0034 3059a

A TEXT-BOOK OF
QUANTIT ATIVE
INORGA-NIC ANALYSIS
THEORY AND PRACTICE
BY
ARTHUR L VOGEL, D.Se. (LOND.); D.Le., F.I.C.

HEAD OF CHE}IISTRY DEPARTMENT. \VOOLWICH POLYTECHNIC:
LATELy' BElT SCIE:NTIFIC RESE;ARCH FELLOW OF THE
IMPERIAL COLLEGE
With Plates and Diagrams
LONGM~NS: GREEN AND CO.

LONDON • NEW YORK- • TORONTO
• 4
By the, same al'tnor
A TEXT BOOK OF
QUALITATIVE CHEMICAL ANALYSIS
With Illusttations
" An exceilent guide to the student from
the early stages in his career up to the
Honours courses."-Nature.
~. 't. ~o. 64. l
Imperial Veterinary Res~ .:', "'' '
Library,! '540\(OG
MUKTESW ............__~~____.._______..._...'
Clas's. ~~ 'chem1str,lT.
Register No. 4014. Room No.
Jnward No. 3059. Jl Shelf No.
Received. 25-5-' 43 Book No.
~QIPC-S2-rrr·6.25(PD)-6.6.38-250.
QUANTITATIVE INORGANIC ANALYSIS
LONGMANS, GREEN AND CO . LTD.
OF PATERNOSTER ROW
43 ALBERT DRIVE, LONDON, S.W.19
17 CHITTARAN]AN AVENUE, CALt:UTTA
NICOL ROAD, BOMBAY
36A MOUNT ROAD, MADRAS
LONGMANS, GREEN AND CO.

55 FIFTH AVENUE, NEW YORK
LONGMANS, GREEN AND Cpo
2 I 5 VICTORIA ST.REET, TORONTO
First Published 1939

Reprinted June 1941
and November 1942
CODE NUMBER: 85287.
Printed In Great Britain by

Lowe and Bryd6.ne Printers Limited, London. N.\V.IO
PREFACE
IN ~riting this bot>k, the author had as his primary object the
proyision of a complete up-to-date text book of quantitative'
inorganic analysis, both theory and practice, at a moderate
price to i>meet the requirements of University and College
students of all grades. It is believed that the material con-
tained thereiq. is sufficiently comprehensive to cover the
syllapuses of all exa~inations in which quantitative, inorganic
analy:;is plays a part. The elementary student has been
provided for, and those sections devoted to his needs have
been'" treated in considerable detail.' The volume should
therefore be of value to the §tudent throughout the whole'
of his career. The book will be suitable inter alia for
students preparing for the various Intermediate B.sc. and
Higher School Certificate Examinations, the Ordinary and
Higher National. Certificates in Chemistry, the Honours and
Special B.Sc. of the Universities, the. Associateship of the
Institute of Chemistry, and other Examinations of equivalent
standard. It is. hoped, also, that the wide range of subjects
discussed within its covers will result in the v~lume having'a
special appeal to practising .analytical chemists and to all
those workers in industry-and research who have occasion to
utilise methods of inorganic quantitative analysis.
The kind reception accorded to the author's .Text Book of
Qualitative Chemical Analysts by teachers and' reviewers
seems to indicate that the general. arrangement of that
book has met with approval: The companion volume on
Quantitative Inorganic Analysis tallows essentially similar
lines. Chapter I is devoted to .the theoretical basis of
quantitative inorganic analysis, Chapter II to the experi-
mental technique of quantitative analysis, Chapter III to
volumetric analysis, Chapter IV to gravimetric analysis
(including electro-analysis), Challter V to colorimetric analysis
and Chapter VI to gas analysis; a-comprehensive Appendix
has been added, which contains much useful matter for the
, chemist. .The experimental side is based
practising a.nalytical
vi Preface
esseI1tially upon the writer's experience with large classes of
students of various grades. Most of the determinations have
been tested out in the laboratory in collaboration with the
author's colleagues and senior students, and in some cases
this has resulted in slight modifications of the details given by
the original authors. Particular emphasis has been laid upon
recent developments in experimt-mtaJ technique: Frequently
the source of certain apparatus 'Or chemicals has.been given in
the text: this is not intended to convey tbe impression that
these materials cannot be obtained from other'sources, but
merely- to indicate that the author's own experience is con-
fined to the particular products mentioned.
The ground covered by the book can best be j~dged by
perusal of the Table of Contents. An attempt has been.made
to strike a balance between the classical and modern prgce-
dures, and to present the subject of analytical chemistry as it
is to-day. The theoretical aspect has been stressed thr.ough-
out, and numerous cJ.;oss-references are given to Chapter I
(the theoretical basis of quantitative inorganic an~lysis) ..
No references to the original literature are given in-the text.
This ts because the introduction of such references would have
considerably increased the size and therefore the price of the
book. However, a discussion on the literature of 'analytical
chemistry is given in the Appendix (Section A. 3). With the
aid of the various volumes mentioned therein-which should
-be available in all libraries of analytical chemistry-and the
Collective Indexes of Chemical Abstracts or of British
Chemical Abstracts, little difficulty will, in general, be
experienced in finding the original sources of most of the
determinations described in the book. " . .
In the preparation of this volume, the author has utilised
pertinent material wherever it was to be found. While' it is
impossible to acknowledge every source indi~idually (see, for
example, Section A. 3), mention must, however, be made of
Hillebrand and Lundell's ,Applied Inorganic Analysis (1929)
and of Mitchell and Ward's Modern Methods in Quantitative
Chemical Analysis (1932): In conclusion, the writer wishes
to express his thanks: to Dr. G. H. Jeffery, A.I:C., for
reading the galley proofs and making numerous helpful sug-
gestions; to Mr. A. S. Nic~elson, B.Sc., for reading some
of th~ g~lley proofs; to his laboratory steward, Mr. F. Mathie,
for preparing a numbet of the diagrams. including most
of those in Chapter VI, and for his assistance in other way~;
to Messrs. A. Gallenkamp.and Co., Ltd., oj London, E.C.2, arid
Preface V'll
to Messrs. Fisher Scientific Co., of Pittsburgh, Pa., for provid-

ing a number of diagrams and blocks;* and 'to Mr. F. W.
Clifford, F.L.A., Librarian to the Chemical Society, and his
able assistants for their help in the task of searching the
extensive literature.
A~y suggestions for improving the book will, be gratefully
receIved by the author.
Woolwich Polytechnic, London, S.E.lS.
June 1939.
• Acknowledgement to other firms and individuals is made in the body of
the text.
CONTENTS
CHAPTER I
mE THEORETICAL BASIS OF QUANTITATIVE. CHEMIC~ ...
~ANALYSIS
PAGE
1, 1. Electrolytic dissociation 1
J:, -2. The law of mass action 4
1, 3. Application of the law ot mass action to solutions of
electrolytes , 5
1, 4. Strengths of acids and bases . ij
I, 5. Activity and activity coefficient 8
I, 6. Ionisation of polyba~ic acids . 9~
I, 7. Common ion effect II
I, 8. Solubility product 13
I, .9. Completeness of precipitation. Quantitative effects of a
common ion 17
I, 10. Limitations of the solubility product principle 19
I, 11. Fractional precipitation . . .' 21
I, 12. Complex ions . . . 22
I, 13. Effect of acids upon the solubility of a precipitate 25
I, 14. Effect of .temperature upon the solubility of a precipitate 26
I, 15. Effect of the solvent upon the solubility of a precipitate 26
I, 16. The ionic product of water 26
I, 17. The hydrogen ion exponent, pH 28
1, 18. The hydrolysis of salts . : 30
1, 19. Hydrolysis constant and degree of hydrolysis .- 33
1, 20. Buffer solutions . 38
VOLUMETRIC ANALYSIS
I, 21. Volumetric analysis . . . . . 43
I, 22. Classification of reactions in volulltetric analysis 44
"1, 23. Equivalent weights. Normal solutions . 45
I, 24. Advantages of the use of the equivalent svstem for the
preparation of standard solutipns .- 53
I, 25. Preparation of standard solutions 5U
1, 26. Primary standard substances 56
I, 27. Neutralisation indicators 58
1, 28. Preparation of indicator solutiovs 67
1. 29. Mixed indicators . . . (jS
I, 30. Universal or multiple-range indicators 70
I, 31. Neutralisation curves. . , , , 70
I, 32. Neutralisation of a strong acid and a strong base V
I, 33. Neutralisation of a weak acid with a strong base 7.t
I, 34. Neutralisation of a weak base with a strong acid 77
I, 35. Neutralisation of a weak acid with a weak base, 78
I, 36. Neutralisation of a polybasic acid with a strong base. SO
I, 37. Titration of solutions of hydrolysed s~lts. Displacement
titrations , , . , ., . '&2
I, 3S. Choice of iItdicators in neutralisation reactions 84
ix
x Contents
THEORY OF PRECIPITATION AND COMPLEX-FORMATION
. PROCESSES
PAGE
I. 39. Precipitation reactions. 86
I. 40. COplplex-formation reactiolls 89
I. 41. Determination of end points in precipitation and in
complex-formation reactions 90
1,42. THEORY OF OXIDATION-REDUCTION. REACTIONS 9.6.

I, 43. Electrode potentials 98
1,44. Concentration cells . . . 101
1,45. Calculation of the e.m.f. of a voltaic cell 101
1,46. Oxidation-reduction cells . 102
I, 47. Calculation of the oxidation potential 103
'1,48. Equili"brium constants of oxidation-reduction reactions 104
·1, 49. Change of the electrode potential during the titration of an
oxidant and a reductant. Oxidation-reduction curve 109
1,50. Indicators for the determination of end points in oxidation-
reduction titrations ll2
1,51. Potentiometric titration 116
1,52. Determination of pH . . . . llB.
I, 53. Experime:ntal details concerning potentiometric titrations.
Recent developments in the tec;Imique oLpotentiometric
titration and of other physical methods of a~alysis 129
THEORY OF GRAVIMETRIC ANALYSIS

I, 54. Gravimetric analysis 136
I, 55. Precipitation methods . 136
I. 56. The colloidal state 140
I, 57. Supersaturation and precipitate formation 145
I, 58. The purity of the precipitate. Cbprecipitation 147
I, 59. Conditions of precipitation 149
I, 60. Washing of the precipitate 150
I, 61. Fractional precipitatiort' 152
I. 62. Organic precipitants . 158
I. 63. Volatilisation or evolution methods 169
ELECTRO-ANALYTICAL METHODS
I, 64. Theory of electro-analysis 171
I, 65. Decomposition potentials 172
I, 66. Electrode reactions 171>
I, 67. Overvoltage 176
I. 68. Completeness of deposition 177
I.. 69. Electrolytic separation of metals 179
r: 70. Character of the deposit . ISO
I. 71. Separation by miscellaneous physical methods. lSI
I. 72. Errors in quantitative analysis IS5
I. 73. Classification of errors lS7
I. 74. Minimisation of errors. . . 190
r. 7S. Significant figures and ~omputations 192
Contents Xl
CHAPTER II
EXPERIMENTAL TECHNIQUE OF QUANTITATIVE ANALYSIS
PAGE
n, 1. Balance, weights, and weighing 194-
n, 2. Description of a typical analytical balance 195
n. 3. The requirements of a good balance 197
n, 4. Weights. Reference masses 198
n. 5. Care and use of the balance 200
n. 6. Methods of weighing 201
n, 7. Errors in weighing 208
n. 8. Calibration of weights 212
n, 9. Other types of balances 216
GENERAL APPARATUS, REAGENTS AND OPERATIONS

n; 10. General apparatus 218
n, 11. Reagents . 226
n, 12. Purificatio~ Qf substances 228
n. 13. Heating apparatus 230
n. ~4. Sampling of solids . 233
n, 15. Crushing and grinding. 234-
n, 16. Solution of the sample 235
TECHNIQUE OF VOLUMETRIC ANALYSIS

n, 17. Unit of volume. .• 238
fi, 18. Standard temperature 239
n, 19. Volumetric apparatus 239
n, 20. Measuring flasks 24()
n, 21. Pipettes . 243
n, 22. Burettes . 246
n, 23. Weight burettes 253
n, 24. Graduated cylinders . 254
n, 25. Storage and preservation. of solutions in volumetric
analysis. ' 254
TECHNIQUE OF GRAVIMETRIC ANALYSIS

D, 26. Precipitation 258
n, 27. Filtration . 259
D, 28. Filter papers . 259
D, 29. Macerated filter paper . 262
D, 30. Filter mats. Gooch crucibles 263
n. 31. Preparation of a Gooch crucible 26a
n, 32. Munroe crucibles . . . . . 266
n, 33. Crucibles fitted with pertnanent porous plates > • 266
n, 34. Washing of precipitates . . . . 267
D, 35. Technique of filtration . . 269
n, 36. Drying and ignition of precipitates 270
D, 37. The care and use of platinum vessels 274
n, 38. Perforated screens for cruci.bles . 277
xii Contents
CHAPTER III
VOLUMETRIC ANALYSIS
ACIDIMETRY AND ALKALIMETRY
PAGE
m, 1. Preparation of a standard acid. . . . . 278
m, 2. Preparation of constant boiling point hydrochloric acid 279
m. 3. Direct preparation of O.lN hydrochloric acid from the
constant boiling point acid 280
m, 4. Preparation of O.lN hydrochloric acid ann standardisa-
tion 281
m. 5. Preparation of standard alkali 286
Ill, 6. Standardisation of approx. O.lN sodium hydroxide 289
m,' 7. Other standard substances for acidimetry and alkali-
metry 291
m. 8. Standgrd barium hydroxide (baryta) solution 295
m, 9. D. * of the Na 2C0 3 content of washing soda 295
In, 10. D. of the strength of glacial acetic acid 297
m. 11. D. of sulphuric acid in the concentrated acid 297
m, 12. D. of sulphuric acid in the fuming acid 298
m, 13. D. of carbon dioxide in the atmosphere 298
m, 14. D. of a mixture of carbonate and hydroxide. Analy~is
of commercial caustic soda 299
m. 15. D. of a mixture of carbonate and bicarbornlte 301
m. 16. D.' of phosphoric acid in commercial orthophosphoric
acid 302
m. 17. D. of boric acid and b o r a x , 303
m. 18. D. of ammonia in an ammonium salt. 305
ill, 19. D. of nitrates (sodium nitrate in Chile saltpetre) 306
m. ·20. D. of nitrogen by Kjeldahl's methoa. 307
PRECIPITATION ~ND COMPLEX FORMATION REAC1'IONS

m, 21. General d~scussion 310
m. 22. Preparation of O.lN silver nitrate 310
m. 23. Standardisation of the silver nitrat~ solution 311
m. 24. D. of chlorides 313
m. 25. D. of bromides 313
m. 26. D. of chlorides and bromides with standard mercurous
perchlorate solution 315
m, 27. D. of iodides . 316
m, 28. D. of thiocyanates. . . 317
m. 29. D. of mixtures of haltdes with absorption indicators 317
m. 30. D. of mixtures of halides by an indirect method 317-
m, 31. Preparation and use of O.IN ammonium or potassium
thiocyanate. Titrations according to Volhard's
method 318
m. 32: D. of silver in a silver alloy 320
m. 33. D. of chlorides (Volhard's method) 321
m. 34. D. of bromides (Volhard's method) 322
m. 35. D. of iodides (Volbard's method) 322
*D. = Determination.
Contents Xlll
PAGE
m. 36. D. of hyposulphites 322
m. 37. D. of cobalt 323
m. 38. D. of nickel 324
m. 39. D. of fluoride as lead chlorofluoride 324
m. 40: D. of arsenic as silver arsenate . 326
m. 41. D. of cyanides '!327
m. 42. D. of nickel b)' potassium cyanide 328
m. 43. D. of thiocyanates 330
m. 44. D. of barium and of sulphates . 332
OXIDATION-REDUCTION TITRATIONS. OXIDIMETRY AND

REDUCTIMETRY
m. 45. General discussion 334
OXIDATIONS WITH POTASSIUM PERl'vIANGANATE

m. 46. Discussion 334
m. 47. Preparation of O.lN potassium permanganate 338
ill, 48. Standardisation- of permanganate solutions. 339
m. 49. Permanence of potassium permanganate solutions 343
m. 50. D. of ferrous iron 344
m. 51. Reduction of ferric to ferrous iron 344
III,.- 52. D. of iron in ferric ammonium sulphate 351
m. 53. D. of total iron in an iron ore 352
m. 54. D. of calcium in calcium carbonate 352
m. 55. Analysis o~ hydrogen peroxide . 354
ill, 56. Analysis of sodium peroxide . 355
m. 57. D. of manganese dioxide in pyrolusite 355
m. 58. D. of nitrites . 357
m. 59. D: of persulphates ... 358
m. 60. D. of manganese in steel (bismuthate method) 358
m. 61. D. of formates an,d of formic acid 361
m. 62. D. of selenium 362
OXIDATIONS WITH POTASSIUM DICHROMATE
m. 63. General discussion 364
m. 64. Preparation of 0.1N potassium dichromate 366
m. 65. Standardisation of potassium dichromate solution
against iron 367
m. 66. D. of ferrous iron .. 368
m, 67. D. of ferric iron (iron in ferric ammonium sulphate) 369
m. 68. D. of total iron in an iron ore .. 369
m. 69. D. of ferrous and ferric iron in an iron ore . 370
m. 70. D. of chromium in a chromic salt 370
m. 71. D. of chromium in chromite 371
m. 72. D. of manganese in steel or in manganese 'ore (Pattin-
son's methoq) 373
OXIDATIONS WITH CERIC SULPHATE

m, 73. General discussion 375
m, 74. Preparation of 0.1N eerie sulphate 378
m. 75. Standardisation of eerie sulphate solution 379
XIV Contents
PAGE"
m, .76.D. of oxalates 382
m, 77. D. of iron in an iron ore 383
ill, 78. D. of ferrocyanides 383-
m, 79. D. of nitrites. . 384
m, 80. D. of tellurium
"
385
m, 81. D. of thallium. 385
ill, 82. D. of cerium 386
.
OXIDATIONS WITH MANGANIC SULPHATE
m, 83. General discussion ..... 387
m. 84. Preparation and standardisation of manganic sulphate
solutions 387
m, 85. Applications of manganic sulphate solutions 388
REDUCTIOXS WITH TITANOUS SALTS

m, 86. Genetpl discussion .. 389
m, 87. Preparation of titanous solutions 390
m, 88. Standardisation of titanous solutions 391
m, 89• .D. of iron in an ore 392
REDUCTIONS WITH AMALGAMATED ZINC AND WITH LIQUID

AMALGAMS
m, 91. D. of uranium. 394
m. 92. D. of titanium 394
ru, 93. D. of molybdenum 395
m, 94. D. of vanadium 395
ru, 95. D, of titanium (zinc amalgam method) 396'
m. 96. D. of tungsten 398
m, 97. Reductions with chromous salts 399

m, 98. Reductions with vanadous salts 400.
OXIDATION AND REDUCTION PROCESSES INVOLVING lODINE
IODIMETRY AND IODOMETRY

ru. 99. General discussion . . 401
m, 100. ~etection of the end point , . 403
ru,101. Preparation of O.IN sodium thiosulphate 405
m. 102. Standardisation of sodium thiosulphate solutions 407
m.103. Preparation of O.IN iodine 414
m, 104. Standardisation of iodine solutions . 415
m. 105. D. of -copper in -crystallised copper sulphate 418
m. 106. D. of copper in an ore 419
m.107. D. of manganese dioxide in pyrolusite 421
m, 108. D. of cblorates and of bromates 423
m.109. Volumetric D. of lead 424
m, 110. Analysis of hydrogen peroxide. ., 425
m, 111. D.d the avaibi.ble chlorine in bleaching powder 426
Contents xv
PAGE
m, 112. D. of hypochlorites . 428
m, 113.- D. of antimony 428
m, 114. D. of antimony in antimonic oxide and in antimonates 429
m,115. :B. of arsenic in arsenic oxide and in arsenates 430
m, 116. D. of tin •. . 430
m, 117. D. of antimony and tin in type metal 432
m,118. D. of sulphurous acid and of sulphites 433
m,119. D. of hydrogen sulphides and of sulphides 434
m, 120. D. of ferricyanides . ' . . . 435
m, 121. D. of ferric iron (iodometric method) . 436
OXIDATIONS WITH POTASSIUM IODATE

m, 122. General discussion .... 428
m,l23. Preparation of O.IN (0.025M)'potassium iodate 440
m,l24. D. of iodides . 441
m,l25. D. of arsenic or of antimony 442
m,126. D. of copper. . . 442
m,l27. D. of mercury 443
m, 128.D. of tin 444
m,129. D. of peroxides (lead, barium and manganese dioxides) 445
m, 130.D. of hydrogen peroxide . " 446
m, 131.D. of hydrazine 447
m,132. D. of thallium. .' 447
m,133. D. of iodides by ceric sulphate 447
m,lM. D. of vanadium 448
OXIDATIONS WITH POTASSIUM BROMATE

m,136. Preparation of O.IN potassium bromate 451
m,137. D. of antimony or of arsenic .... 451
m, 138. D. of metals by means of 8-hydroxyquinoline (" oxine") 452
m .. 139. D. of hydroxylamine . . • . . . 458
OXIDATIONS WITH CHLORAMINE-T

m, 140. General discussion. Preparation of ()o.IN solution 459
m, 141. D. of antimony 460
m, lA2. D. of nitrites . 460
m, 143. D. of tin . 461
m, 144. D. of ferrocyanides 462
.. MISCELLANEOUS VOLUMETRIC DETERMINATIONS

m,145. D. of zinc with standard potassium ferrocyanide
solution 463
m,l46. D. of phosphorus '. 464
m. 147. D. of temporary and permanent hardness of water 466
m. 148. D. of sodium . 472
m,149. D. of potassium 472
m,l50. D. of cadmium 472
xvi Contents
CHAPTER IV
GRAVIMETRIC ANALYSIS
PAGE
IV, 1. General discussion 473
IV. 2. Note book entries, and calculations 473
IV, 3. Calculations of gravimetric analysis. Chemical factors 473
SIMPLE qRAVIMETRIC DETERM,INATIONS

IV, -4. D. of watex; of hydration in crystallised barium chloride 474
IV, 5. D. of chloride as silver chloride 475
IV, 6. D. of sulphate as barium sulphate 478
IY, 7. D. of sulphur in iron pyrites 482
IV, 8. D. of iron as ferric oxide 484
IV, 9. D. of aluminiqm as aluminium oxide . 488
IV, 10. D. of calCium as oxalate, carbonate or oxide 490
IV,I1. D. of magnesium as pyrophosphate 494
IV,I2. D. of nickel as the dimethylglyoxime complex ", 497
SYSTEMATIC GRAVIMETRIC ANALYSIS

IV, 13. General discussion 500
IV. 14. Lead 501
IV, 15.
IV, 16.
Silver
Mercury
_. 503
504
IV. 17. Bismuth 506
IV,18. Cadmium 508
IV, 19. Copper 512
IV. '20. Arsenic' 515
IV. 21. Antimony 517
IV,22. Tin 519
IV, 23. Molybdenum ,_ . 522
IV, 24. Selenium and tellurium .525
IV,25. Platinum 527
IV, 26. Palladium . 529
IV, 27. Gold 531
IV, 28. Aluminium 533
IV, 29. Beryllium 536
IV, 30. Chromium 539
"
·IV,31. Iron ." 542
IV, 32. Nickel 545
IV, 33. Cobalt 546
IV,34. Zinc 549
IV. 35. Manganese .. 553
IV, 36. Vanadium 555
IV, 37. Uranium . 556
IV,38. Thorium . 557
IV,39. Cerium 558
IV, 40. Titanium- 559
IV, 41 .. Zirconium. 561
IV. 42. Thallium 563
IV, 43. Calcium 564
Contents xvii
PAGE
IV. 44. Strontium . 564
IV. 45. Barium 565
IV. 46. Magnesium 567
IV. 47. Sodium 567
IV. 48. Potassium 570
IV. 49. Lithium 574.
IV. 50. Ammonium 574-
IV. 51. Tungsten 575
IV. 52. Chloride 576
IV. 53. Bromiqe 566
IV. 54. Iodide 576
IV. 55. Thioc-yanate 577
IV. 56. Cyanide 578
IV. 57. Fluoride "578
IV. 58. Chlorate 580
IV. 59. Perchlorate 581
IV. 60. Iodate 582
-IV. 61. Sulphate . 582
IV. 62. Sulphide 583
IV. 63. Sulphite 583
IV. 64. Thiosulphate 584
IV. 65. Phosphate 584
IV. 66. Phosphite 586
IV. 67. Hypophosphite 587
IV. 68. Oxalate •. 587
IV. 69. Borate 588
IV. 70. Silicate 588
IV. 71. Fluosilicate 591
IV. 72. Ferrocyanide .' 591
IV, 73. Ferri~yanide 592;
IV. 74. Nitrite 592
IV. 75. Nitrate 592
IV. 76. Carbonate . 592
ELECTROLYTIC DETERMINATIONS
IV, 77. Apparatus 596
IV, 78. Copper 606
IV. 79. Lead 610
IV. 80. Cadmium 611
IV, 81. Nickel 612
IV, 82• .,Cobalt 61:l
IV. 83. Silver 614.
IV, 84. Zinc . .. ..... 614.
IV. 85. Electrolytic separation and determination of copper and
nickel. . . .'. . . . • 615
IV. 86. SIMPLE GRAVIMETRIC SEPARATIONS 617
ANALYSIS OF COMPLEX MATERIALS
IV. 87. Analysis of brass 625
IV. 88. Analysis of bronze . . . . 631
IV. 89. Analysis of German silver (nickel silver) 634
KVlll Contents
.l'A(iE
iV, 90. Analysis of solder . 636
lV, 91. Analysis of a silver-coinage alloy 638
·IV,92. Analysis of steel 639
IV, 93. Analysis of an aluminium alloy . 652
IV, 94.
lV, 95.
Analysis of a limestone or dolomite
Analysis of 'a felspar .
• 654
661
IV, 96. Analysis of portland cement .' 666
CHAPTER V
COLORIMETRIC ANALYSIS
V, 1. General discussion .. 670
V, 2. Theory of spectrophotometry and colorimetry . 761
V, 3. Classification 'of m.ethods of "colour" measurement or
comparison 675
V, 4. A. Standard series method . 678
B. Duplication method 68'2
D. Balancing method . 683
E. Gradation or step-photometer method 687
F. Photo-electric photometer method 689
V, 5. Turbidimetry and nephelometry.. . . 692
V. 6. D. of th(l. pH of solutions by colorimetric methods . 693
V, 7. Some general remarks upon colorimetric determinations 697
V, 8. D. of ammonia ". 699
V, 9. D. of nitrites 700
V, 10. D. of iron 701.
V,11. Manganese 703
V, 12. Titanium 705
V,13. Vanadium 708
V, 14. Chromium 709
V, 15. Aluminium 710
V, 16. Molybdenum 711
V, 17. Nickel 713
V, 18. Cobalt 714
V.19. Arsenic 714
V, 20. Antimony 721
V,21. Tin 723
V,22. Lead 724
V,23. Copper • 726
V.24. Bismuth 727
V, 25. Silver 728
V. 26. Silicate . 729
'V,27. D. of other elements and radicals 739
CHAPTER VI
GAS 4NALYSIs
VI, 1. General discussion 732
VI, 2. Sampling 733
VI, 3. Purification of mercury 734
Contents xix
PAGE
VI, 4. Correction of volume of gas for temperature and pressure 736
VI, 5. Calibration of gas measuring vessels 739
VI, 6. Apparatus employed in- gas analysis 740
VI, 7. Absorbents for the various gases . 762
VI, 8. D. by explosion and combustion methods 767
VI, 9. Exercises in gas analysis 770
GAS-VOLUMETRIC METHODS OF ANALYSIS FOR SOLIDS AND

LIQUIDS
VI, 10. The Lunge nitrometer 771
VI, 11. D. of nitrates 772
VI. 12. D. of hydrogen peroxide 774
VI. 13. Evaluation of pyrolusite 775
VI. 14. D. of available clj.lorine in bleaching powder 776
VI. 15. Evaluation of zinc dust 776
VI. 16. D. of ammonium salts 776
APPENDIX
A, 1. International atomic weights, 1939 778
A. 2. Chemical factors 779
A, 3. The literature of analytical chemistry 783
A, 4. Densities of acids at 20°C 788
A, 5. Densities of alkaline solutions at 20°C 790
A, 6. Data on the strength of aqueous solutions of the common
acids and ammonium hydroxide 790
A, 7. Samples and solutions for elementary analysis 791
A, 8. Analysed samples 792
A, 9. Analytical reagents 798
A, 13. Buffer solutions . 808
A, 11. Comparison of metric and other units 813
A,12. Decimals of an inch, 1.S.W.G., and millimetres 815
A,13. The Greek alphabet. . . . 815
A, 14. Suggested scheme of study for Intermediate Science and
Engineering, Higher School Certificate, and Ordinary
National Certificates in Chemistry, and Examinations of
similar character 816
A, 15. Suggested scheme of study for RSc. Special Chemistry,
Honours B.Sc., Chemistry, Associateship of the Institute
of Chemistry, and Examinations of similar scope 817
A, 16. Four-figure logarithms 822
At 17. Five-figure logarithms. 824
Index 843
LIST OF PLATES
To face page
PLATE 1. Cambridge pH meter, electrical circuit Of latter,
and electrodes for use with this instrument 131
PLATE II. R D. H. Lovibond nesslerimeter 682
PLATE III. Bausch and Lomb Duboscq colorimeter 684
PLATE IV. "Spekker" photo-electric absorptiometer 689
CHAPTER I
THE THEORETICAL BASIS OF QUANTITATivE

CHEMICAL ANALYSIS
I, 1. Electrolytic diSsociation.-Many of t.he reactions of
qualitative and quantitative analysis take place in aqueous
solution. It is therefore necessary to have a general know-
ledge of the conditions which exist in such solutions. It is
assumed that the reader is familiar with the broad concepts
of Arrhenius's theory of ,electrolytic dissociation. *
Ionisation of acids and bases in solution.-An acid may be
defined as a substance which, when dissolved in water, under-
goes dissociation with the formation of hydrogen ions as the
only positive ion:
HCl ~ H+ + Cl- ;
.HNO s ~ H+ + NO s -.
Actually the free hydrogen ions H + (or protons) do not exist
in the free st~te in aqueous solution; each hydrogen ion
combines with one molecule of water to form the hydroxonium
ion HaO+. The hydroxonium ion is a hydrated proton. The
above equations are therefore more accurately written:
.. '
HCI + H 2 0 ~ HaO+ + Cl-;
HNO a + H 20 ~ HaO+ + NOs - .
The ionisation may be attributed to the great tendency of the
free hydrogen ions H+ or protons to combine with water
molecules to form hydroxonium ions. Hydrochloric and
nitric acids are almost completely dissociated in aqueous
solution in accordance with the above equations; this is
readily detected by freezing point measurements and by other
methods.
Polybasic acids ionise in stages. In sulphuric acid, the
first hydrogen atom is almost completely ionised:
H 2SO, ~ H+ + HSO, - ,
or H 2SO, + H 2 0 ~ HaO+ + HSO,-.
• 'A short account will be found in the author's Text Book of Qualitati",
Chemical Analysis, 1937 (Longmans, Green and Co,).
B 1
2 Quantitat~ve Inorganic Analysis
The second hydrogen ~tom is only partially ionised, except in
very dilute solution:
HS0 4 - ~ H+ + S04--'
or HS0 + H 20 ~ HaO-f' + SO(--.
4-
Phosphoric acid also ionises in stages :

H aP0 4 ~H+ + H 2P0 4- ~H+ + HP0 4-- ~H+ + P0 4 - - - ;
or H aP0 4 + HP ~ HaO+ + H 2 P0 4 - , •
H 2P0 4 - + H 2 0 ~ H30+ + HP0 4 - - ,
HP0 4 --" + H 2 0 ~ HaO+ + PO" ---.
The successive stages of ionisation are known as the primary,
secondary and tertiary ionisatibps respectively. As already
mentioned, these do not take place to the same degree. The
primary ionisation is. always greater than the secondary, and
the secondary very much greater than the tertiary.
Acids of the type of acetic acid H.C2Ha02 give an almost
normal freezing point depression in aqueous sol~tion; th,e
extent of dissociation is accordingly small. It is usual,
therefore, to distinguish between acids which are completely
or almost completely ionised in solution and those which are
only slightly ionised. The former are termed strong acids
(examples: hydrochloric, hydrobromic, hydriodic, iodic,
nitric and perchloric acids, primary ionisation of sulphuric
acid) and the latter are called weak acids (examples:
nitrous acid, acetic acid, carbonic acid, boric acid, phosphor-
ous acid, phosphoric acid, hydrocyanic acid, and lfydrogen
sulphide). There is, however, no sharp division between the
two classes. Methods for the measurement of the relative'
strengths of acids are described in Section I, 4.
A base may be defined as a substance which, when dis-
solved iI1 water, undergoes dissociation with the formation
of hydroxyl ions OH- as the only negative ions·. Thus
sodium hydroxide, potassium hydroxide and the hydroxides
of the alkaline earth metals a:J;e almost completely dissociated
in aqueous solution:
NaOH ~ Na+ + OH-;
Ba(OHh ~ Ba++ + 20H-.
• A more general eoncept of acids and bases is due to Bronsted; this is now
generally accepted. He defines an acid as a substance which ionises to yield
hydrogen ions or prbtons, and a base as a substance which combines with
hydrogen ions. An acid is accordingly a proton donor and a base a protor
acceptor. Both can be defined by the expression:
A ~H+ + B.
where A is the acid and B is the base.
The Theoretical Basis of Quantitative Analysis 3
These are strong bases. Ammonium hydroxide solution,
however, is 'a weak base. Only a small concentration of
hydroxyl ions is produced in aqueous solution:
NH 40H ~ NH4
,
+ + OH-,
or
Salts. The structure pf numerous salts in the solid state
has been investigated by means of X-rays and by other
methods, and it has been shown that they are composed of
charged atoms or groups of atoms held together in a crystal
lattice. When thes,e salts are dissolved in a solvent of high
dielectric constant such as water, or are hectted to the melting
point, the orystal forces are weakened and the substances
dissociate into the pre-existing charged particles Of" ions, so
that the resultant liquids ar,e good conductDrs of electricity.
The complete dissociation of salts in aqueous solution is the
modern view and is now almost universally accepted. There
are, however, some exceptions, and these have, in some cases,
been supported by X-ray measurements. Feebly ionised
salts (weak electrolytes) are exemplified by the cyanides,
thiocyanates and halides of mercury and cadmium, and by
lead acetate.
The theoretical implications of the theory of complete
ionisation, due to Debye, Hiickel and Onsager, have been
fully worked out by these authors. In particular, they have
been able to account for the increasing equivalent conductance
with decreasing concentration over the concentration range
O-O·002N. For full details the reader must be referred to
text-books of physical ,chemistry ; a short account will be
found in the author's Text Book of Qualitative Chemical
Analysis.
It is important to realise that whilst complete ionisation
occurs with strong electrolytes, this does not mean that the
effective concentrations of the ions are the same at all con-
centrations, for if this were the case, the osmotic properties
of aqueous solutions could not be accounted for. The varia-
tion of osmotic properties is ascribed to changes of the
" activity of the ions; these are dependent upon the elec-
11
trical forces between the ions. Expressions for the varia!ions

of the activity" or of related quantities, applicable to dilute
II
solutions, have also been deduced by the Debye-Hiickel-

Onsager theory. Further consideration of the conception of
" activity will be found in Section I, 5.
11
4 Quantitative Inorganic Analysis
I, 2. THE LAW OF MASS ACTION
Guldberg and Waage in 1'867 clearly state<:]. the law of mass
action (sometimes termed the law of chemical equilibrium)
in the form: the velocity of a <:hemical reaction is propor-
tional to the product of the active masses of the reacting
substances. For the present we shall. interpret "active
mass" by concentration and express it in gram molecules
(or mols) per litre. By applying the law to homogeneous
systems, i.e., to systems in which all the reacting molecules
are present in one phase, for example in solution, we can
arrive at a mathematical expression for the condition of
equilibrium in a reversible reaction.
Let us consider first the simple reversible reaction at
constant temperature:
A+B ~C +D.
The velocity with which A and B react is proportional to their
concentrations, or
VI = kl X rA] X [B],
where ki is a constant known as the velocity coefficient, and
the square lJrackets in he<1vy type denote the molecular
concentrations of the substances enclosed within. the brackets.
Similarly, the velocity with which the reverse reaction occurs
is given by:
v2 = k2 X [C] X [D].
At equilibrium, the velocities of the reverse and the forward
reactions will be equal (the equilibrium is a dynamic and not a
static one) and tHerefore VI = v2 ,
or kl X [A] X [B] = R2 X [C] X [DJ,
or [C] X [D] = kll = K.
[A] X [B] kl
K is the equilibrium: constant of the reaction at the given
temperature. .
The expression may be generalised. For a reversible
reaction represented by: -.
PIAl + P2 A 2 + PaAa + .... ~ q1B 1 + q2 BlI + q3 B a + .... ,
where PI' P2' P3 and q}l Q2' qa are the number of molecules of
reacting substances, the condition for equilibrium is given by
the expression :
[Bllql
K.
• The Theoretical Basis 0/ Quantitative Analysis 5
This result may be expressed in words: When equilibrium is
reached in a reversible reaction, at constant temperature, the
product, of the molecular concentrations of the resultants
(the substances on the right hand side of the equation)
divided by the product of the molecular concentn!.tions of the
reactants (the substances on the left·hand side of the equation),
each concentration being raised to a power equal to the num-
. ber of molecules of that substance taking part in the reaction,
is constant.
It must be pointed out that there is no relation between the
velocity at which .a reaction attains equilibrium and the
equilibrium constant. The velocity of a reaction increases
rapidly with rise of temperature, and is also affected by
certain substances. Those substances which increase the
rate of reaction are termed positive catalysts, whilst those
which decrease it are called negative catalysts.
I, 3. Application of the Jaw of mass action to solutions of
electrolytes.-Weak (or slightly dissociated) electrolytes, such
as acetic. acid and ammonium hydroxide, undergo reversible
dissociation when dissolved in water, The equilibrium be-
tween the undissociated molecules and the ions f-or such
electrolytes can be investigated by the law of mass action.
The law cannot be applied to strong electrolytes, such as
salts, since dissociation is complete, or virtually complete.
The equilibrium 'which exists in a dilute sol~tion of acetic
acid at constant temperature is :
H.C2Ha02 ~ H+ + C2 H a0 2- *.
Applying the law of mass action, we have:
[C 2H a0 2-] X [H+] / [H.C2 H a0 2] = K.
K is the ionisation constant or dissociation constant at constant
temperature. The term affinity constant is 'sometimes em-
ployed for acids and bases. If one gram equivalent of the
electrolyte is dissolved in V litres of solution (V = lie, where
c is the concentration in gram equivalents per litre), and if a
is the degree of ionisation at equilibrium, then the amount of
unionised electrolyte will be (1 - a) gram equivalents, ~nd
the amount of each of the ions will be a gram equivalents.
The concentration (gram equivalents per litre) of unionised
• Strictly speaking the equilibrium should be written:
H.C,HaC. + H 20 .p H.O+ + C.H.O.-
no free protons (H+) exist in aqueous solution. The older form will. however,
be retained for the sake of simplicity.
acetic acid will therefore be (1 - a);V, and the concentration
of each of the ions a/V. Substituting in the equilibrium
equation, we obtain the expression:
a2 / (1 - a)V = K or a 2c / (1 - a) = K.
This is known as Ostwald's dilution law. The agreement of
the" la'Y" with experiment is illustrated by the following
results for acetic acid (Table I).
TABLE I. EQUIVALENT CoNDUCTANCE AND DISSOCIATION CoNSTANT OF
ACETIC ACID AT 25°C.
Concn x 104 A, a K X 105

1·873 102·5 0·264 1·78
5'160 65·95 0·170 1·76
9·400 50·60 0·130' 1·83
24·78 31·94 0·080 1·82
38'86 25·78 0·066 1·83
56·74 21·48 0·055 1·84
,
68·71 19·58 0·050 1·84
92'16 16-99- 0·044 1·84
112·2 15·41 0.040 1·84
0 388·6
The mean classicai or Ostwald dissociation constant of"acetic

acid at 25°C is 1·82 X 10-5 , Similar results-the individual
variations may sometimes be slightly greater-are obtained
for other weak electrolytes.
It 4. Strengths of acids and bases.-It has already been

stated (Section It 1) that the properties of acids are the pro-
pertie~ of the hydrogen ion H+ (or, more correctly, the
hydroxonium ion HaO+). For any given total concentration
of acid, the concentration of hydrogen ions will depend upon
the degree of dissociation a; the strength of an acid will thus
depend .upon the value 6f a at a given concentration. The
dissociation constant gives a relationship between a and the
concentration, and it accordingly also represents a measure
of the strength of the acid or a measure of its tendency to
undergo dissociation.
The properties of bases, according to the roost elementary
view, depend upon the hydroxyl ion* and the ionisation
* See footnote, Section I, 1.
The Theore!ical Basis of Quantitative Analysis 7 .
TABLE II. DISSOCIATION' CoNSTANTS AT 25 oc.
Acid. Formula K. pK. = - log K.

-
Acetic H.C.H.O. 1·82 X 10-· 4·75
, 1·76 X 10- 5 t 4·75
Benzoic H.C 7 H.O. 6·37 X 1O-s t 4·20
o-Chlorobenzoic H.C,H.ozCI 1·20 X 1'0-· t 2·92
Formic H.CHO z 1·77 X 10-' t 3·75
Furoic
Hydrocyanic
Nitrous
. H.C.HaO.
H.CN
H.NO.
6·78 X 10- 5
7·2 X 10-10
4·6 X 10- 4
t 4·17
9·14
3·33
o-Nitrobenzoic H.C 7 H.O •. NO z 6·00 X 10-' t :l·22
Phenylacetic
Adipic . -
H,CsH 70.
H •. C.HsO,
4·88.x 10- 5
}(13'72 X 10- 5
t
t
4·31
4·43
K. 3·86 X 10-" t 5·41
Carbonic H •. CO. K 1 4·31 X 10-' t 6·37
K. 5·61 X 1O-11 t 10·25
Hydrogen H •. S K 1 9·1 X 10-' 7·04
Sulphide K.l·2 x.I0-IS 14·92
Malonic Hz.c.H.O, Kl 1·40 X 10-3 t 2·85
K. 2·20 x 10-" t 5·66
Oxalic H •. CzO, K 1 5'9 X 10-' t 1·23
K.6·4 X 10-· t 4·19
Phthalic H •. CsH,O, K l l·2 X 10-' 2·92
K.l X 10-· 5'00
Succinic H 2 .C,H,O, KI 6·63 X 10-· t 4·18
K. 2·54 X 10- 8 t 5·60
Sulphuric Hz·SO, K. 1·15 X 10-2 t 1·94
Sulphurous Hz·SO. K l l'7 X 10-' 1·77
K z 1·0 X 10-7 7·00
d-Tartaric H •. C,H,O. K 1 1·04 X 10-3 t 2·98
K z 4·55 X 10- 5 t 4·34
Boric H.H.BO. Kl 5·80 X 10-10 t 9·24
Citric H •. C.H.O, Kl 9·20 X 10-' t 3·04
K. 2·69 X 10- 5 t 4·57
K. 1·34 X 10- 6 t 5·87
Phosphoric H •. PO.
. K I 7'52 X 10-'
K. 6·23 X 10-'
t
t
2·12
7·21
K.5 X 10-13 t 12·30
Base K. pK. = - log K
Ammonium NH,OH }·8 X 10 S 4·74
hydroxide 1·79 X 10-5 t 4·75
Aniline C.HsNH. * 4·0 X 10-10 9·40
Ethylamine C.H.NH. * 4·6 X 10-' 3·34
Dimethylamine (CH.)zNH * 5·20 X 10-' t 3·28
Methylamine CH.NH. * 4·38 X 10-' t 3·36
Piperidine C.HllN * 1·3 X 10-" 2·88
Quinoline CoH,N * 6·0 X 10-10 9·22
Trimethylamine (CH')3 N * K.3·1 5·45 X 10-' t 3·26
Calcium Ca(OH). X 10-' t 1'51
hydroxide
•
Barium Ba(OH), K.2·3 X 10-1 t 0·64
hydroxide
* All compounds marked with an asterisk are of the ammonium type, e.g.,
aniline is really anilinium hydroxide:
(C.N.NH.)OH ~ (C.H.NHa)++ OH-.
t Figures marked with a t are the true or thermodynamic dissociation
constants. (See Section I, 5.)
8 Qltantitative Inorganic Analysis
constant will likewise be a measure of the strength of the base.
For very weak or slightly ionised electrolytes, the expression
0. /(1 -
2 o.)V = K reduces to 0. 2 = KV or a. = V1{V, since 0.
may be neglected in comparison with unity. Rence for any
two weak acids or bases at a given dilution V (in litres), we
have 0.1 = VK;V and 0. 2 = vK 2 V, or 0.1/0.2 ~ vKdvK2'
Expressed in words, for any two weak or slightly dissociated
electrolytes at equal dilutions, the degrees of dissociation are
proportional to the square 'roots of their ionisation constants._
Some values for the dissociation constants at 25°C jor weak
acids and bases are collected in Table II.
I, 5. Activity and activity coefficient.-In our deduction
of the law of mass action it was assumed that the effective
concentrations or active masses of the components could be
expressed by the stoichiometric concentrations. According
to modern thermodynamics, this is not strictly true. The
rigid equilibrium equation for, say, a binary electrolyte:
AB ~ A+ + B-,
is aA+ X aB- / aAB = Ka,
where aA+, aB- and aAB represent the activities of A+, B-
and AB respectively, and Ka is the true or thermodynamic
dissociation constant. The concept of acti,vity. a thermody-
namic quantity, is due to G. N. Lewis. The quantity is
related to the concentration by a factor, termed the activity
coefficient :
activity = concentration X activity coefficient.
Thus at any molar concentration
aA+ = lA+' [A+], aB- =lB-' [B-], and aAB = lAB' [AB],
where 1 refers to the activity coefficients, and the square
brackets to the molar concentrations. Substituting in the
above equation, we obtain:
1 A+' [A+] xl B-' [B-] = [A+].[B-] X f A+ XfB- = K
fAB . [A B) [AB] lAB a'
This is the rigorously correct expression for the law of mass

action as applied to weak electrolytes.
The activity coefficient varies with the concentration. For
ions, it varies with the valency, and is the same for all dilute
solutions having the same ionic.. strength, the latter being a
measure of the electrical field existing in the solution. The
term ionic strength, designated by the symbol !-'-' was intrQ-
The Theoretical Bast's of Quantitative Analysis 9
duced by Lewis and Randall in 1921, and is defined as equal

to one half of the sum of the products of the concentration of
each ion ml)ltiplied by the square of its valency, or f-I- =
0·5.Ec,~, where c. is the ionic concentration in gram mols per
litre of solution and z, is the val_ency of the ion concerned.
An example will make this clear. The ionic strength of 0,]
molar HNO a solution containing 0·2 molar Ba(NO a)2 is given
by:
0'5{ 0·1 (for H+) + 0·1 (for NO a-) + 0·2 X 22 (for Ba++)
+ 0·2 X 2 (for NO a-) }
= 0·5 {l'4} = 0·7.
The activity coefficient depen~s upon the total ionic strength

of the solution in a manner whic'll is discussed in Section I, 10.
Generally speaking, it is comparatively difficult to determine
activity coefficients experim~ntally, partic~larly in concen-
trated solutions and in mixtures of ions of several valency
types. The activity coefficients of unionised mole«ules do
not differ considerably from unity. For weak electrolytes in
which the ionic concentration and therefore the ionic strength
is small, the error introduced by neglecting the difference
between the actual values of the activity coefficients of the
ions, fA+ and fB-" and unity is small « 5 per cent.). Hence
for weak electrolytes, the true or therplOdynamic expression
reduces to [A+] X [B-] / [AB] = K; the constants obtained
by the use of simple concentrations will be accurate to 2-6
per cent. Such values are sufficiently precise for purposes 0f
calculation in quantitative analysis. It must, however, be
pointed out that precisioR values for the dissociation constants
of weak electrolytes 'can be obtained by the use of special
methods; the discussion of these is outside the scope of this
volume.
Strong electrolytes will be assumed to be completely
dissociated, and no correction for activity coefficients will be
made for 'dilute solutions.
I, 6. Ionisation of polybasic acids.-When a polybasic
acid is dissolved in water, the various hydrogen atoms undergo
ionisation to different extents. For a dibasic acid H2A, the
primary and secondary ionisations can be represented by the
equations:
H2A ~H+ + HA- (1)
HA- ~ H+ + A-- (2).
B
If the dibasic acid is a weak electrolyte, the 1aw of mass action
may be applied, and the following expressions obtained:
[H+] X [HA-] / [H2A] = KI (IA),
[H+] X [A--] I [HA -] = K2 (2A).
K I.and K 2 are known as the primary and 'SeCOndary dissociation
constants respectively. Each stage of the dissociation
process has its own ionisation c~mstant, and the magnitudes
of these constants give a measure of the extent to which each
ionisation has proceeded at any given concentration. The
greater the value of Kl relative to K 2 , the smaller will be the
secondary di~sociation, and the. greater must be the dilution
before the latter becomes appreciabte. It is therefore possible
that a dibasic (or polybasic) acid may behave, so far as
dissociation is concerned, as a monobasic acid. This is
indeed characteci.stic of many polybasic acids (see Table II). '
A tribasic acid H3A (e.g., orthophosphoric acid) will simi-
larly yield three dissociation constants, Kv K2 and K 3 ,
which may be computed in an analogous manner:
H3A ~ H+ + H 2A- (3) ;
H 2A ...· ~ H+ + HA-- (4);
HA-- ~ H+ + A--- (5).
We can now apply some of the theoretical considerations to
actual examples encountered in analysis .
. Example 1. To calculate the concentrations of HS- and S~- in a
saturated solution of hydrogen sulphide.
A saturated aqueous solution of hydrogen sulphide at 25°C, at to
atmospheric pressure, is approximately O'~ 1 molar. The primary aqd
secondary dissociation constants are 9·1 X lO-8 and 1·2 X 10-15
respectively.
Thus [H+] X [HS-] / [H 2S] = 9·1 X lV-a (i),
and [H+] X [S--] / [HS-] = 1·2 X 10-15 (ii).
The very much smaller value of. K2 indicates that the secondary
dissociation and consequently [S--] , is exceedingly small. It follows,
therefore, that only the primary ionisation is of importance, and
[H+] and [HS-] are pra.ctically equal in value. Substituting in
equation (i) :
[H+] = [HS-] and [H 2S] = 0·1, we obtain
[H+l = [HS-] = y!9·1 X 10-8 X 0·1 = 9·5 X 10-6•
Both the eqUilibrium equations must be satisfied simultaneously;
by substitution of these values for [H+] and [HS-] in equation (ii),
we obtain 9·5 X IO-a X [S--] =,1,2 X 10-16 X 9·5 X 10-5, or
[5--1 = 1·2 X 10-15 , which is the value for K 2 •
If we multiply equations (i) and (ii) together and t~anspose :
[S--] = 1·1 X 1O-23 /[H+p.
Thus the concentration of the sulphide ion is inversely proportional
to the square of the hydrogen ion concentration, i.e., if we, say,
double [H+] by.the addition of a strong acid, the [S--] would be
reduced to 1/22. or 1/4 of its original vahfe. .
i,7. Common ion effect.-The concentration of.a particu-
lar ion in an ionic reaction can be increased by the addition
of a compound which produces that ion upon dissociation.
The particular ion is thus derived from the compound already
in solution aItd from the added reagent, hence the name
common ion. We shall confine our attention to the case in
which the original compound is a weak -electrolyte in order
that the law of mass action may be applicable. The result is
usually that there is a higher concentration of" this ion in
solution than that derived from the original compound alone, •
and new equilibrium' conditions will be produced. Examples
of the calculation of the common ion effect are given below.
In general, it may be stated. that if the total concentration of
the common ion is only slightly greater than that which the
original compound alone would furnish, the effect is small ;
if, however, 'the concentration of the common ion is very
much increased (e.g., by the addition of a completely disso-
ciated salt), the effect is very great and may be of considerable
practical importance. Indeed, the common ion effect pro-
vides a valuable method for controlling the concentration ,of
the ions furnished by a weak electrolyte.
Example 2. To calculate the sulphide ion concentration in a 0·25
molar hy,drochloric acid solution saturated with hydrogen sulphide.
This concentration has been chosen since it is that at which the
sulphides of the metals of Group II are precipitated. The total
concentration of hydrogen sulphide will be approximately the same
as in aqueous solution, i.e., O·IM; the [H+] will be equiJ) to that of
the completely dissociated Hel, i.e" 0·25M, but the [S--] will be
reduced below 1·2 X 10-15
Substituting in equations (i) and (ii) (Section I. 6). we find:
9·1 X 10-8 X [H 2S]' 9·1 X 10-8 X 0·1
[HS-] = [H+) 0.25 = 3·6 X 10-8 ,
__ 1·2 X 10-15 X [HS-] 1·2 X lO-rs X 3·6 X 10-8
[S )= [H+] - 0.25
= 1·7 X 10- 22 •
12 QuantitaHve Inorganic Analysis
Thus by changing the acidity from 9·5 X 10-5M (that present in
saturated H 2S water) to 0·25M, the sulphide ion concentration is
reduced from 1·2 X 10-15 to 1·7 X 10- 22•
Example 3. What effect has the addition of 0·1 gram molecule
(8·20 g.) of anhydrous sodium acetate to I litre of 0·1 molar acetic
acid upon the degree" of dissociation of the acid?
The dissociation constant of acetic acid at 25°C is 1·82 X 10-5•
The degree of ionisation a in O·IM solution (e = 0'1) may be com-
puted by solving the quadratic equation:
[H+] X [C 2H a0 2-] = ~ = 1.82 X 10-5.
[H.C 2H a0 2] (I - a)
For our purpose it is suffiGiently accurate to neglect a in (I - q.)
since a is small: -
.'. a = "'I/Ke - V~I-'8~2-X-'-1-0--4 = 0·0135.
Hence in O'IM acetic acid,
[H+] = 0-00135, [C 2H a0 2-] = 0'00135, and [H.C 2H a0 2] = 0·0986.
The concentrations of sodium and acetate ions produced by the
addition of the completely dissociated sodium acetate are:
[Na+] = 0·1, and [C 2H a0 2-] = 0·1 gram molecule respectively.
The acetate ions from the salt will tend to decrease the ionisation of
the acetic acid since K is constant, and consequently the acetate ion
concentration derived from it. Hence w,e may write [C 2H a0 2-] =
0·1 for the solution, and if a' is the new degree of ionisation, [H+] =
a'e = O'la', and [H.C 2H a0 2] = (1 - a')e = 0-1, since a' is negli-
gibly small.
Substituting in the mass action equation:
[H~ X {C 2H a0 2-] = O·la' X 0·1 = 1.82 X 10-5,
[H.C 2H 30 2] 0·1
or a' = 1·8 X 10-4.
[H+] = a'e = 1·8 X 10-5•
The addition of a tenth of an equivalent weight of sodium acetate
to a 0·1 molar solution of acetic acid has decreased the degree of
ionisation from 1'35% to 0 ·018%, and the hydrogen ion concentration
from 0·0135 to 0·000018.
I Example 4. What effect has the addition of 0·5 gram molecule

(26-75 g.) of ammonium chloride to I litre of 0·1 molar ammonium
hydroxide solution. upon the degree 'of dissociation of the base?
.
(Dissociation constant of NH40H = 1·8 X 10-5.)
In O·IM ammonia solution a = Vl·8 X 10-1i X 0·1 = 0·013.
Hence [OH-] = 0'0013, [NH,+] = 0'0013, and [NH,OH] = 0·0987.
Let a' be the degree of ionisation in the presence of the added
ammonium chloride. Then [OH-] = a'e = 0·10.', and [NHPH] =
(1 - a')e = 0·1, since a' may be taken as negligibly small. The
addition of the completely ionised ammonium chloride will of
necessity decrease the [NH,+] derived from the base and increase
[NH"OH], since K is constant under all conditions, Now [NH,,+] =
0·5, as a first approximation.
Substituting in the equation:
[NH/] X [OH-] _ 0·5 X 0·10.' = 1'~ X 10-5 ,
[NH"OH] 0·1
or a' -= 3·6 X 10-5 and [OH-] = 3·6 X 10.-6 •
The addition of half an equivalent weight of ammonium
chloride to a 0·1 molar s6lution of ammonium hydroxide has
decreased the degree of. ionisation from 1'35% to 0'0036%,
and the hydroxyl ion concentration from 0·0013 to 0'0000036.
1, 8. Solubility product.-For sparingly soluble §alts (i.e.)
those of which the solubility is less than 0·01 g-ram molecules
per litre) it is an experimental fact that the product of' the
total molecular concentrations of the ions is a constant at
constant temperature. This product S is termed the
solubility product. For a binary electrolyte:
AB ,.: A+ + B-,
SAB = [A+] X [B-].
In general, for an electrolyte ApBg, which ionises into PA +, ...
and qB- ••.. ions:
ApBg": pA.+' ... + qB-' ... ,
SAliBI = [A+' .. ']P X [B- ... .]g,
A plausible deduction of the sol~bility product relation is
the following. When exc~ss of a sparingly soluble electrolyte,
,say silver chloride, is shaken up with water, some of it passes
into solution to form a saturated solution of the salt and the
reaction appears to cease. The following equilibrium is
.actually present (the silver chloride is completely ionised in
solution) :
AgCl (solid) ,.: Ag+ + CC
The velocity of the forward reaction depends only upon the
temperature, and at any given temperature:
VI = kr'
where kI is a constant. The velocity of the reverse reaction is
proportional to the concentrations of each of the reacting
substances; hence at any given temperature:
V2 = k2 X [Ag+] X [cq,
where K'1, is another constant. At equilibrium the two
velocities are equal,
i.e., kl = k2 X [Ag+] X [Cl-],
or [Ag+] X [Cl-] = kl/k2 = SAgel'
A more rigid deduction of the solubility product pril}ciple
can be obtained along the following.lines. For simplicity, we
shall again consider a binary electrolyte:
AB ~ A+ + B-.
The solid is completely dissociated in dilute solution; no
undissociated molecules will therefore be present. It can be
shown by thermodynamics that in.a saturated solution of any
salt containing varying amounts of a more soluble .salt with a
common ion, the products of the activities of the two ions in
the various saturated solutions are the same:
i.e., aA+ X aB- = constant (activity product).
In the very dilute solutions with which yve are at present
concerned, the activities may be taken as practically equal to
the concentrations,* so that [A+] X [B-] = constant. .
Some experimental results, due to ]ahn (190f), illustrating
the approximate consta"'ncy of the solubility product of silver
chloride in the presence of varying concentrations of chloride
ion (from potassium chloride) are collected in Table III.
TABLE III. EFFECT OF CHLORIDE IONS UPON THE SOLUBILITY PRODUCT
OJ AgCl.
[KCl] I[Cl-] X 10_3 I[Ag+] X 10 8 SAgCl= [~g+].[CqXI010

0·00670 6·4 1·75 1·12
0·00833 7·9 1·39 1·10
0·01114 10·5 1·07 1·12
0·01669 15·5 0·74 1·14
0·03349 30·3 0·39 1·14
* Activity = activity coefficient X concentration, i.e., a = fr.. The activity

product may therefore be written:
fA+[A+] X jB-[B-) = constant.
In very dilute solution j is proportional to (1 - k -vc). and if c is small, this
quantity is approximately unity.
It is important to note that tb,e solubility product relation
applies with sufficient accuracy for purposes of quantitative
analysis only to saturated solutions of slightly soluble elec-
trolytes and with small additions of other salts. In the
presence 6f.large concentrations of salts, the ionic concentra-
tion, and therefore the ionic strength of the solution, will
increase. This will, in general, lower the activity coefficients
of both ions, and consequently the ionic concentrations (and
therefore the solubility) must increase in order to maintain the
solubility product constant (see Section I, 10 for a more
detailed discussion).
The great importance of the conception of solubility product
lies in its bearing upon precipitation from solution, which is,
of course, one of the principal operations of quantitative
analysis .. The solubility product is the ultimate value which
is attained by the ionic product when equilibrium has been
established between the solid phase of a difficultly soluble salt
and the solution. If the experimental conditions :j.re such
that the ionic product is different from the solubility product,
then the system will attempt to adjust itself in such a manner
that the ionic and solubility products are equal in value.
Thus, if, for a given electrolyte, the product of the concen-
trations of the ions in solution is arbitrarily made to exceed
the solubility product, as, for example, .by the addition of a
salt with a common ion, the adjustment of the system to
equilibrium results in the precipitation of the solid salt,
provided supersaturation conditions are excluqed. If the
ionic product is less than the solubility product or can
arbitrarily be made so, as, for example, by complex salt
formation or by the formation of weak electrolytes, then a
furth()r quantity of solute can pass into solution until the
solubility product is attained, or,. if this is not possible, until
all the solute has dissolved.
Table IV contains the solubility products (S) a.t the labora-
tory temperature (ca. 20°C) of some common sparingly soluble
substances. Values for pS = -loglo S are given in the
last column. The student is referred to text-books of
physical chemistry for a· description of the methods for
determining solubility products. .It must, however, be
pointed out that the various methods do not always yield
consistent results, and what appear to be the best representa-
tive figures 'are given in the Table.
TABLE IV. SOLUBILITY PRODUCTS AT THE LABORATORY
TEMPERATURE.
Substance Solubility pS Substance Solubility pS

Product Product
-
1'5 X 10-11 -18·82
AgBr 3'5 X 10-11 12·46 FeS
AgCNS 7·1 X 1O-1a 12·15 Hg,Br. 1·3 X 10-18 17·89
AgC1 1·2 X 10-10 9·92 Hg,C1 a 2·0 X 10-11 20'70
Ag,CrO, 1·7 X 10-12 11'77 Hg,I. 1·2 X 10-28 27·92
AgI 1·7 X 10-18 15·77 HgS 4 X IO- A 52·4
Ag.PO, 1·8 X 10-18 17:74 KH.C,H,O. 3·0 X 10-' 3·52
Ag,S 1·6 X 10-" 48·80 K 2 PtCl. 4·9 X 10- 5 4·31
AI(OH). 8·5 X 10-28 22·07 MgCO a 2·6 X 10- 6 4'58
BaCO. 1·9 X 10-· 8·72 MgC.o, 8·6 X 10- 6 4·07
BaC,O, 1·7 X 10-7 6·77 MgF. 7·0 X 10-' 8·15
BaCrO, 2·3 X 10-10 • 9·64 Mg(NH,)PO. 2·5 X 10-18 12·60
BaSO, 1·2 X 10-1 • 9·92 ,Mg(OR), 1·5 X 10-11 10·82
CaCO. 1·7 X 10-8 7·77 Mn(OR), 4·0 X 10-1< 13·40
CaC,O, 3·8 X 10- 9 8·42 MnS 1·4 X 10-16 14'85
Ca.C.R,O. 7·7 X lO-7 6·11 Ni(OH). 8·7 X 10-19 18·06
(tartrate) NiS 1·4 x-10-2& 23·85
CaF. 3·2 X 10-11 10·49 PbBr. 7·9 X 10- 5 4·10
CaSO, 2·3 X lO-' 3·64 PbCl. 2·4 X ~o-' 3·62
CdS 3·6 X lO-" 28·44 PbCO. 1·7 X 10- 11
10'77
Co(OR)s 1·6 X 10-18 17·80 PbCrO, 1·8 X 10-1& 13·74
CoS 1·9 X 10-'7 26·72 PbF. 7·0 X lO-1 8·15
Cr(OR)a 2·9 X lO-2I 28·54 PbI. -1'4 X 10-8 7·85
Cu,(CNS), 1'7 X 10-11 10·77 PbS 4·2 X 10- 29
27·38
Cu,Br, 4·1 X 10-8 7·39 PbSO, 2·3 X ~0-8 7·64
Cu,Cl. 1·4 X 10-' 5·85 srCO. 4·6 X 10-8 8·34
Cu,I, 2·6 X 10-12 11·58 srC.O, 1·4 X 10-"- 6·85
CuS 8·5 X lO-46 44'07 SrSO, 3·6 X 10- 7
6·44
Fe(OR). 1~6 X 10-1& 13·80 Zn(OR). 1·0 X 10-18 18·00
Fe(OH). 1·1 X lO-88 35·96 ZnS ca. 1 X 10-zo 20·0
A few examples may help the reader to fully understand the

subject. The concentrations are expressed in gram mols,per
litre for the calculation of solubility products.
Example 5. The solubility of silver chloride is 0·0015 grams per

litre. Calculate the solubility product.
The molecular weight of silver chloride is 143·5. The solubility is
therefore 0·0015/143'5 = 1·05 X 10-5 mols per litre. In a satur-·
ated solution the dissociation, AgCI ~ Ag+ + CC is complete; 1
mol of AgCI will give 1 mol each of Ag+ and Cl-. Hence [Ag+] =
1·05 X 10-5 and [cq'= 1·05 X 10:-5 •
SAgCl = [Ag+] X. [cq = (1'05 X 10-5) X (1'05 X 10-5)
= 1·1 X 10-10• '
Example 6. Calculate the solubility product of silver chromate,

given that its sqlubility is 2·5 X 10- 2 grams per litre.
Ag2Cr0 4 ~ 2Ag+ + Cr0 4--.
The molecular weight of Ag 2Cr04 is 332, hence the solubility
= 2·5 X 10 -2 /332 = 7·5 X 10 -5 mols per litre.
Now 1 mol of Ag2Cr04 gives two mols of Ag+ and 1 mol of Cr04--,
therefore SAg2CrO( = [Ag+p X [Cr0 4--] •
= (2 X 7·5 X 10-5)2 X (7·5 X 10-5)

= 1·7 X 10-12 •
Example '7. The solubility product of magnesium hydroxide is

3·4 X 10-11. Calculate its solubility in grams per litre.
Mg(OH)2 (solid) ~ Mg++ + 20H-.
[Mg++] X [OH-P = 3·4 X iO-11.
The molecular weight of magnesium hydroxide is 58. Each mol of
magnesium hydroxide, when dissolved, yields one mol of magnesium
ions and two mols of hydroxyl ions. If the solubility is -x gram
mols per litre, [Mg+j] = x and [OH-] = 2x. Substituting these
values in the solubility product expression:
x X (2X)2 = 3·4: X 10-11
or x = 2·0 10-4 g. mols per litre
X
= 2·0 X 10-4 X 58 = 1·2 X 10- s·g. per litre.
I, 9. Completeness of precipitation. Quantitative effects

of a conu,non ~on.-An important' application of the solubility
product principle is to the calculation of the solubility of
sparingly soluble salts in solution of salts with a common ion.
Thus the solubility of a salt MA in the presence of a relatively
large amount of the common M+ ions,* supplied by a second
salt MB, follows from the definition of solubility product:
[M+] X [A-] = SMA
or [A-] = SMA I [M+] (i).
The solubility of the salt is represented by the [A-] which it

furnishes in solution. It is. clear that the addition of a com-
'mon ion will decrease the solubility of
the salt.
, "
Example 8. Calculate the solubility of silver chloride in O·OOIM
and O'OIM sodium chloride solutions n:spectively (SAgCI = 1·1 X
10-10).
In a saturated solution of silver chloride
[cq = 1·1 ·x 10-10 = 1·05 X 10-5 g. m'oI per litre; this may be
neglected in comparison with the excess of Cl- ions added.
• This ena~les us to neglect the concentration of M+ ions supplied by the
sparingly soluble salt itself, and thus to simplify the calculati?n.
For [cq = = 1·1 X 10-10/1 X 10-3
1 X 10- 3, [Ag+]
= 1·1 X 10-7 g.
mol per litre.
For[Cq = 1 X 10-', [Ag+] = 1·1 X 10-1 °/1 X 10- 2
• = 1·1 X 10-8 g. mol per litre.
Thus the solubility is decreased 100 times in O'OOlM'sodium chloride
and 1000 times in O'OlM sodium chloride. SImilar results are ....
obtained for O'OOlM and O'OlM silver nitrate solution.
Example 9. Calculate the solubilities of silver chromate in
O'OOIM and O·OlM silver nitrate solutions, and in O'OOlM and O'OIM
potassium chromate solutions (Ag2Cr0 4 : S.P. = 1·7 X 10-12,
solubility = 7·5 X 10-5 mols per litre).
[Ag+]2 X [Cr0 4--] = 1·7 X 10-12,
or [CrO,--] = 1·7 X 10-12 / [Ag+)2.
For O·OOlM silver nitrate solution: [Ag+] = 1 X 10-3 ,
.'. [Cr04 --] = 1·7 X 10-12 /1 X 10-6 = 1·7 X 10-6 mols perlitre.
For 0'01M silver nitrate solution:
[Cr0 4- - ] = 1·7 X 10-12 /1 X 10-4 = 1·7 X 10- 8 mols per litre
The solubility product equation gives:
1 '-=-7-x--'1=-=0---'1-;;2'-;/-;;'[C""'r"""'O""',===-].
[Ag+] = y'C:-
For [CrO,--] = 0'001, [Ag+] = y'1'7 X 10",,12 /1 X 10-3
= 4·1 X 10-5 mols per litre.
For [Cr0 4- - ] = 0·01, [Ag+] = y''-:;-I---::'7'-x~I'''''0---;1=2-'/~1-x"-710''---':2
= 1·3 X 10-5 mols per litre.
This decrease in solubility by the common ion effect is of
fundamental importance in gravimetric analysis. By the
addition of a suitable excess of a precipitating agent, the
solubility of a precipitate is usually decreased to so small a
value that the loss by washing is negligible. Let us consider
a specific case-the determination of silver as silver- chloride.
Here the chloride solution is added to the solution of the silver
salt. If an exactly equivalent amount is added, the resultant
saturated solution, of silver chloride will contain 0·0015 g.
perlitre (Example 5). If 0·2 g. of silver chloride is produced
and the volume of the solution and washings is 500 ml, the
loss, owing to solubility, will be 0·00075 g. or 0·33 per cent.
of the weight of the salt; the analysis would then be 0·33 per
cent. too low. By using an excess of the precipitant, say, to a
concentration of O·OIM, the solubility of the silver chloride is
reduced to 1·5 X 10-5 g. per litre (Example 8), and the loss
will be 1·5 X 10-Ii X.0·5 X 100/0'2 = 0·0033 per cent.
Spver chloride is therefore very suitable for the quantitative
determination of silver with high a~curacy.
1,10. Limitations of the solubility product principle.-for a
binary electrolyte MA, thermodynamical considerations lead
to the result that the product of the activities of the two ions
is constant at constant temperature: .
i.e., aM+ X aA- = constant (activity product = f5 MA ) (1).
Now activity = concentration X activity coefficient, or a =.
cJ. Substituting in (1), we obtain:
[M+]fM+ X [A-]fA- = fSMA (2).
in which fM+ and fA- denote the activity coefficients of M+
and A-respectively. For a saturated solution of a sparingly
soluble salt in water, the ionic concentration is so small'that
!M+ and fA- are not far removed from unity p.nd the expres-
sion (1) reduces to the classical solubility product equation.
Numerous experiments
80 .....- - - - - - - - - - ,
have been carried out to
., MI(NO ,. determine the
s
accuracy
£: 70 KNO, with which the'solubility
x
product principle applies.
Some results of Bray and
Winninghoff (1911) upon
the solubility of thallous
chloride in the presence
.. "" " . SO'- added·of various .
salts are shown
• (Cak.' m FIg. 1-1. The d6tted
line represents the effect
of a simple uni-univalent
" , Ag+ added
salt with a common ion,
100~~O~.O::-:2~O~,04~O~.':"06~0~.0::-:8~0010 (Calc.) calculated by assuming
Normolitv of added salt complete dissociation of
FIG. ,I·I.-Influence of various salts
the salt. The salts with
on the solubility of TICI. a common ion decrease
the solubility to nearly
the same extent, and are in approximate agreement with the
values computed on the basis of the simple solubility product
principle. Fig. 1-2 shows graphically some selected results for
silver sulphate. The addition, in moderate concentrations,
of a common univalent ion, like Ag+, is in approximate
agreement with the simple theory, but no such agreement is
obtained by the addition of a common divalent ion, like
50 40- - , A striking fact that emerges from these studies is,
.that the solubility increases in most cases upon the addition
of electrolytes with no common ion; we may term this influ-
ence the salt effect. Furt,hermore, it was found that the salt
..,
~
x ~o
~
;::;:
~8
..§' TlNO.
~
~ 4 ._. " BaCI 2
~ Calc: •••• _ •• : ••••• ____ • I
00 0.02 0·04 0·06 0'06 0·10 0·12 0·14 0·16

Normaldy 0/ added salt
FIG. 1-2.-Influence of various salts on the solubility of Ag.SO,.
effect depends upon the valency of the ions of the precipitate

and upon the nature of the added electrolyte; usually it
increases markedly with increasing valency of the ions of the
precipitate. The addition of a salt with no common ion will.
increase the ionic strength of the solution and hence decrease
the activity coefficients of both ions of the sparingly soluble
salt; the solubility of the latter must increase in order that
the .activity product, [M+]. [A-] . fM+ . fA-, may be kept
constant. It can be shown on the basis of the Debye-
Huckel-Onsager theory that for aqueous solutions at 25°C:
ft 0·505z,2· °.5
log f,. - - ----.......:..---
•- 1 + 3'3 X 107a. ft°'s '
where!. is the activity coefficient of the ion, z. is the valency
of the ion concerned, p. is the ionic strength of the solution
(Section I, 5), and a is the average" effective diameter" of all
the ions in the solution. For very dilute solutions (ft 0.5 < 0·1)
the second term. of the denominator is negligible and the
equation reduces to :
log!. = ~ 0·505 Z,2. ft °.5.
For more concentrated solutions (ft°' s > 0'3) a.n additional

term Bp. is added to the equation; B is an empirical constant.
For a more detailed treatment of the influence of salts upon
The Theoretical Basis of Q1tantitative Analysis 21
solubility and solubility product, particularly of the work of
Bronsted and of La Mer, the reader. is referred to text-books
of electrochemistry. * I
It will be clear from the above short discussion that two

factors may come into play wpen a solution of a salt containing
a common ion is added to a saturated solution of a slightly
soluble salt. At moderate concentrations of the added salt,
the solubility will generally decrease, but with higher con-
centrations when the ionic strength of the solution increases
considerably and the activity coefficients of the ions decrease,
t]J.e solubility may actually increase. This is one of the
reasons, provided complex ion formation. is absent, why a
very large excess of the precipitating agent is avoided in
quantitative analysis.
:t 11. Fractional precipitation.-We have thus far con-
sidered the solubility product principle in connexion with the
precipitation of one sparingly soluble salt. We shall now
extend our studies to the case where two slightly soluble salts
may be formed. For simplicity, we shall study the situation
which arises when a precipitating agent is added to a solution
containing two anions, both of which form slightly soluble
salts with the sam~ cation, e.g., when silver nitrate solution is
added to a solution containing both chloride and iodide ions.
The questions which arise are: which salt will be precipitated
first, and how completely' will the first salt be precipitated
before the second ion begins to react with the reagent?
The solubility products of silver chloride and silver iodide
are respectively 1·2 X 10-10 and 1·7 X 10.;-16;
i.e., [Ag+] X [cq - 1·2 X 10-10 (i),
[Ag+] X [r] = 1·7 X 10-16 (ii).
It is evident that silver iodide being less soluble will be
precipitated first since its solubility proauct will be first
exceeded. Silver chloride will be precipitated when the Ag+
ion concentration is greater than
SAgCI 1·2 X '10-10
[Cr] = [CI ]
and then both salts will be precipitated simultaneously.
When silver chloride commences to precipitate, silver ions
* See, for example, S. Glasstone, The Electrochemistry of Solutions. 1937.
p. 127 (Methuen and Co.).
will be in equilibrium with both salts and equations (i) and (ii)
will be simultapeously satisfied, or
A +] _ SAgI _ SAgel
[ g - [r) - [Cl-) (iii) ,
[rl SAgI 1·7 X 10-18 1
and [CI-l = SAgel - 1·2 X 10-10 - 7:1 X 107 (iv).
Hence when the concentration of the iodide ion is about one
millionth part of the chloride ion concentration, silver chloride
will be precipitated. If the initial concentration of both
chloride and iodide ions is O·lN, then silver chloride will be
precipitated when
[r] = 0·1/7·1X10 7 = 1·4x10-7N = 1·SxlO-6 g. per litre.
Thus' an ahn.ost complete separation is theoretically possible.
The separation is feasible in practice if the point at which the
iodide precipitation is complete can be detected. This may
be done: (a) by the use of an absorption indicator (see
Section 1, 41C), or (b) by a potentiometric method with a silver
electrode (see Section I, 53).
For a mixture of bromide and iodide:
[C] SAgi 1·7 X 10-18 1
===---p
[Br-] SAgBr 3·5 X 10-13 2·0 X lOS
Precipitation of silver bromide will Occur when the concentra-
tion of the 'bromide ion in the solution is 2·0 X 103 times that
of the iodide concentration. The separation is therefore not
quite so complete as in the case of chloride and iodide, but
can nevertheless be effected with fair accuracy with the aid
of adsorption indicators (Sections I, 41C and m, 29).
1, 12. Complex ions.-The' increase in solubility of a pre-
cipitate upon the addition of excess of.the precipitating agent
is frequently due to the formation of a complex ion. A
complex ion is formed by the union of a simple ion with either
other ions of opposite charge or with neutral molecules. Let
us examine a few examples in detail.
When putassium cyanide solution is added to a solution of
silver nitrate, a white precipitate of silver cyanide is first
formed because the solubility product of silver cyanide,
[Ag+] X [CN-] = SAgeN (i),
is exceeded. The reaction is expressed:
K+ + CN- + Ag+ + NO a- = AgCN + ~+ + N0 3-,
or CN- + Ag+ = AgCN.
The precipitate dissolves upon the addition of excess of
potassium cyanide, the complex ion [Ag(CN)2r being pro-
duced:
AgCN (s01id) + CN- (excess) ~ [Ag(CN)2r*
(or AgCN + KCN = K[Ag(CN)2]-a soluble complex salt).
This complex ion -dissociates to give silver ions, since the
addition of sulphide ions yields a precipitate of silver s)llphide
(solubility product 1·6 X 10-49), and also silver is precipi-
tated from the complex cyanide solution upon electrolysis.
The complex ion thus dissociates in accordance with the
equation:
[Ag(CNhr ~ Ag+ + 2CN- (ii).
By applying the law of mass action to (ii), we obtain the
dissociation constant or instability constant of the complex
ion:
[Ag+] X [CN-P _ K (iii) .
. £{Ag(CN)d I -
The constant has a value of 1·0 X 10-21 at the ordinary
temperature. By inspection of this expression, and bearing
in mind that excess of cyanide ion is present, it should be
evident that the silver ion concentration must be very small,
so small in fact that the solubility product of silver cyanide
is not exceeded.
Consider now a somewhat different type' of complex 'ion
formation, viz., the production of a complex ion with con-
stituents other than the common ion present in the solution.
This is exemplified by the solu""ility of silver chloride in
ammonia solution. The reactions are: .
Ag+ + Cl- ~ AgCl ;
AgCl (solid) ~ Ag+ + Cl- + 2NHa ~ [Ag(NHa)2]+ + Cl-
(or AgCI + 2NHa = [Ag(NH 3 hJct). '
Here again, for reasons similar to those already given, silver
ions are pI~sent in solution~ The dissociation of the complex
ion is represented by :
[Ag(NHah]+ ~ Ag+ + 2NH a,
and the instability const~nt is given by :
[Ag+] X [NHaP _ _. -8
[fAg( NH a)2}+] , - K - 6 8 X 10 .
• It is usual to employ square brackets to include the whole of a complex
ion. In order to avoid confusion with concentrations for which square brackets .
are also widely used, concentrations will be represented in heavy type and
complex ions either in lighter type or between curly brackets.
24 Qttantitative I norgant'c Analyst's
The magnitude of the instabiiity constant clearly shows that
only a very small silver ion concentration is produced by the
dissociation of the complex ion.
The stability of complex ions varies within very wide limits.
It is quantitatively expressed by means of the dissociation
or instability constant. The more stable the complex, the
smaller is the instability constant, i.e., the smaller is the
tendency of the complex ion to dissociate into its constituent
ions: When the complex ion is very stable, e.g., the ferro-
cyanide ion [Fe(CN)6r---, the ordinary ionic reactions of the
components are not shown. A few selected values of these
constants at the ordinary temperature, determined by
methods involving potentiometric titr~tion (Sectiqns 1,51-53),
are collected iIi Table V; this is instructive and also useful for
reference purposes.
TABLE V. INSTABILiTY CONSTANTS OF COMPLEX IONS
[Ag+J X [NH3]1 [Cd++] x rCN-]·

[{Ag(NH.l.}+ ]
= 6·8 X lO-8 -"---,--'::"_-=---=-
[{Cd[CNl.}- ]
= 1·4 X lO-17
[ Ag +J X [5 •0 3 --J3 = 1.0 lO-18 [Hg++] X [(,;:'-]4

= 6·0 X lO-17
[{Ag(5.0.l.}---J X ({HgCl,} ]
[Ag+] X [CN-J 1 [Hg++] x [r]'
= 1·0 X lO-21 = 5·0 X lO-11
.[{Ag(CNltn [{HgI.} ]
[Cu+J x [CN-J' [Hg++] x [CN-]'
= 5·0 X lO-28
[{Cu(CNJJ ] [{Hg(CNl.} ] = 4·0 x lO-(I
The application of complex ion formation in chemical separa-

tions depends upon the fact that one component may be
transformed into a complex ion which is no longer precipitable
with the precipitating agent, whereas another component is
precipitated. One example may be mentioned here. This
is concerned with the _separatioh of cadmium and copper.
Excess of potassium cyanide solution is added to the solution
containing the two salts when the complex ions [(Cd(CN)J--
and [CU(CN)4r-- respectively ar~ formed. Upon passing
hydrogen sulphide into the solution containing excess of
CN- ions, a precipitate of cadmium sulphide is produced.
Despite the higher solubility product of CdS _(3'6 X 10- 28
as against 8·5 X 10-45 for copper sulphide), the former is
precipitated because the complex cadmium cyanide ion
has a greater instability constant (see Table V).
I, 13. Effect of acids upon the solubility of a precipitate.-
For sparingly soluble salts of a strong acid the effect of the
addition of an acid will b~ similar to that of any other indiffer-
ent electrolyte (compare salt effect in Section I, 10). If the
sparingly soluble salt MA is the salt of a weak acid HA, then
acids will, in general, have a solvent effect upon it. Let us
suppose that hydrochloric acid is added to an aqueous sus-
pension of such a salt. The following equilibrium will be
established:
M+ + A- + H+ + Cl- ~ HA + M+ + CC.
lf the dissociation constant of the acid HA is very small, the
anion A- will be removed from the solution to form the undis-
sociateq acid HA. Consequently more of the salt will pass
into solution to replace the anions removed in this way and
this process will continue until equilibrium is es1:ablished (i.e.,
until [M-t] X [A-] has become equal to the solubility product
of MA) or, if sufficient hydrochloric acid is present, until the
sparingly soluble salt has dissolved completely. Similar
reasoning may be applied to salts of polybasic acids, such as
phosphoric acid (KI = 7·5 X 10-3 ; K2 = 6·2 X 10-8 ;
Ka = 5 X 10-13), oxalic acid (Kl = 5·9 X 10-2 ; K2 =
6·4 X 10-5), and arsenic acid. Thus the solubility of, say,
silver orthophosphate is due to the removal of the P9,---
ion as HPO,-- and/or H 2 P0 4 -:
P0 4- - - + H+ ~ HPO,-- ;
HP0 4- - + H+ ~ H 2 PO,-.
With weak acids, such as carbonic (KI = 4·3 X 10-7 ;
Kz = 5·6 X 10-11), sulphurous (Kl = 1·7 X 10-2 ; K z =
1·0 X 10-7), and nitrous (Kl = 4·6 X 10-4 ) acids, an addi-
tional factor contributing to the increased solubility is the
actual disappearance of the acid from solution either spon-
ta.neously or on gentle warming. An explanation is thus
provided for the well-known solubility of the sparingly soluble
sulphites, carbonates, oxalates, phosphates, arsenites, arse-
nates, acetates, cyanides (with the exception of silver cyanide,
which is actually a salt of the strong acid H[Ag(CNhJ),
fluorides, and salts of other organic acids in strong acids.
The sparingly soluble sulphates (e.g., those of barium,
strontium, and lead) also exhibit increased solubility in acids
as a consequence of the weakness of the second ~tage of
ionisation of sulphuric acid (K2 = 1·2 X 10-2 ) ; .tfie first
stage is completely ionised:
Ba++ + S04-- + H+ + Cl- ~ HS0 4- + Ba++ + Cr-.
Since K2 is comparatively large, the solvent effect is relatively
small. However, in the quantitative separation of barium
sulphate, precipitation is usually carried out in slightly acid
solution in order to obtain a more easily filterable precipitate
and to reduce coprecipitation (Section I, 58).
The precipitation of substances within a controlled range of
pH is discussed in Section I, 61B.
I, 14. Effect of temperature upon the solubility of a
precipitate.-The solubility of the precipitates encountered
in quantitative analysis increases with rise of temperature.
With some substances the infruence of temperature is small,
but with others it is quite appreciable. Thus the solubility
of silver chloride at 10° and 100°C is 1·72 and 21·1 mg. per
litre respectively, whilst that of barium sulphate at these two
temperatures.is 2·2 and 3·9 mg. per litre respectively. In
many instances, the common ion effect reduces the solubility
to so small a value that the temperature effect, which ,is
otherwise appreciable, becomes very small. Wherever pos-
sible it is advantageous to filter while the solution is hot;
the rate of filtration is increased as is also the SOlubility of
f9reign substances, thus rendering their removal from the
precipitate more complete. The double phosphates of
ammonium with magnesium, manganese or zinc as well as
lead SUlphate and silver chloride are usually filtered at the
laboratory Femperature to avoid solubility losses.
I, 15. EHect of the solvent upon the solubility of a precipi-
tate.-The solubility of most inorganic compounds is reduced
by the addition of organic solvents, such as methyl, ethyl and
n-propyl alcohols, acetone, etc. For example, the addition
of about 20 per cent. by volume.of ethyl alcohol renders the
solubility of lead sulphate practically negligible, thus permit-
ting quantitative separation. Similarly calcium sulphate
separates quantitatively from 50 per cent. ethyl alcohoL
Other examples of the influence of solvent will be found in
Chapter IV.
I, 16. The ionic product of water.-K@hlrausch and Hey-
dweiller (1894) found that the most highly purified water
that can be obtained possesses a small but definite conduc-
tivity. Water must therefore be ~lightly ionised in accordance
with the equation: H 20 ~ H+ + OH-*.
"
• Strictly speaking the hydrogen ion H+ exists in water as the hydroxonium
ion HaO+. The electrolytic dissociation of water should therefore be written:
2H.O "" HaO+ + OH-.
For the sake of simplicity, the more familiar symbol H+ will be retained.
The Theoretical Basts of Q1tantita!ive Analysis 27
Applying the law of mass action to this equation, we obtain,.
for any given temperature:
aH+ X aOH-
=
[H+J . [OH-]
X
/H+' /OH-
=
t
a constan ,
[HPJ .
. aH20 /HsO
where a"" [xl. andj", refer to the activity, concentration and

activity coefficient of the species x. Since water is only
slightly ionised, the ionic concenttations will be small and
their activity coefficients may be regarded as unity; the
activity of the unionised molecules may also be taken as
unity. The expression thus becomes:
[H+] X [OH-]
[HPJ = a constant .
.
In pure water or in dilute aqueous solutions, the concentration
of the undissociated water may be considered constant.
Hence:
[H+J x· [OH-J = K w ,
where Kw is the ionic product of water. It must be pointed
out that ,the assumption that the activity coefficients of the
ions are unity and that the activity coefficient of water is
constant applies strictly to pure water and to very dilute
solutions (ionic strength < 0'01); in more concentrated
solutions, i.e., in solutions of appreciable ionic strength, the
electrical environment affects the activity coefficients of the
.ions (compare Section I, 10) and also the activity of the
un ionised water. The ionic product of water will then not
be constant, but will depend upon the ionic environment.
It is, however, difficult to determine the" activity coefficients,
except under specially selected conditions, so that in practice
the ionic product K w , although not strictly constant, is
employed.
The ionic product varies with the temperature, but under
ordinary experimental conditions (at about 25°C) its value
may be taken as 1 X 10-14 with concentrations expressed in
gram ions per litre. This is sensibly constant in dilute
aqueous solutions. If the product of [H+J and [OH-l in
aqueous solution momentarily exceeds this value, the excess
ions will immediately combine to form water. Similarly if
the product of the two ionic concentrations is momentarily
less than 10-14, more water molecules will dissociate until the
eqUilibrium value is attained.
28 QuantitaHve Inorganic Analysis
• The hydrogen and hydroxyl ion cqncentrations are equal in
pure water; therefore [H+] = [OH-] = vKw = 10-7 gram
ions per litre at about 25°C. A solution in whiCh the hydro-
gen and hydroxyl ion concentrations are equal is termed an
exactly neutral solution. If [R+1 is greater than 10-7, the
solution is acid, and if less than 10-7 , the solution is alkaline
(or basic). It follows that at ordinary temperatures [OH-]
is greater than 10-7 in alkaline solution and less ..than this
value in acid solution.
In all cases the reaction of the solution can be quantita-
tively expressed by the magnitude of the hydrogen ion (or
hydroxonium ion) concentration, or, less frequently, of the
hydroxyl -ion concentration, since the following simple rela-
tions between [H+1 and [OH-] exist:
[H+] _:_ [O~-]' and [OH-] ='[~:r

The variation oj K", with temperature is shown in Table VI.
TABLE VI. IONIC PRODUCT OF WATER AT VARIOUS TEMPERATURES.
Temp. (0C) I K .. X 1014 Temp. (0C)

I K., X IOU
0° 0·12 35° 2·09

5° 0·19 40° 2··92
10· 0·29 45 Q
4·02
15° 0·45 50· 5·48
20· • 0·68 55° 7·30
25·
30°
1·01
1·47
60·
I 9·62
I, 17. The hydrogen ion exponent, pR.-For many pur-

poses, especially when dealing with small concentrations, it is
cumbersome to express concentrations of hydrogen and
hydroxyl ions in terms of gram equivalents per litre. A very
convenient method was proposed by S. P. L. Sorensen (1909).
He introduced the hydrogen ion exponent pH defined by the
relationships:
pH = - loglo [H+] = loglo 1/[H+], or [H+] = lo-pH.
The quantity pH is thus the logarithm (to the base 10) of the
reciprocal of the hydrogen ion concentration, or is equal to the
logarithm of the hydrogen ion concentration with negative
sign. This method has the advantage that all states of
acidity and ·alkalinity between those of solutions molar (or
normal) with respect to hydrogen and hydroxyl ions can be
expressed by a series of positive numbers between 0 and 14.
Thus a neutral solution with [H+] = 10-7 has a pH of 7; a
solution with molar (or normal) concentration of hydrogen
ions has a pH of 0 (= 10°); and a solution ,molar with
respect to hydroxyl ions has [H+] = K", / [OH-] = 10-14 /10°
= 10-14 , and possesses a pH of 14. A neutral solution is
therefore one in which pH = 7, an acid solution one in which
pH .< 7, and an, alkaline solution one in which PH > 7.
An alternative definition for a neutral solution, applicable to
all tempe!atures, is one in which the hydrogen ion and
hydroxyl ion .(:oncentrations are equal. In an acid solution
the hydrogen ion concentration exceeds the hydroxyl ion
concentration, whilst in an alkaline or basic solution, the
hydroxyl ,ion concentration is greater,
Example 10. (i) Find the pH of a solution of which [H+]

= 4·0 X 10-6•
Log 4·0 = 0·602, hence log 4-0 X 10-6 is 5·602 = 0·602 - 5
= -4:398
(the decimal part or mantissa of the logarithm is alway~ positive).
pH = -log [H+] = - ( - 4·398) = 4·398.
(ii) Find the hydrogen ion concentration corresponding to
:.;H = 5·643.
PH = -log [H+] = 5·643; . '. log [H+] =- 5·643.
This must be written in the usual form containing a negative
characteristic and a positive mantissa:
Jog [H+] = - 5·643 = 6·357.
Referring to a table of ~ntilogarithms, the number corresponding to
the logarithm of 0·357 is 2·28; the number corresponding to the
logarithm 6·357 is accordingly 2·28 X 10-6•
[H+J is therefore 2·28 X 10-6•
(iii) Calculate the pH of a 0·01 molar solution of acetic acid (the
degree of dissociation is 12'5%), .
The hydrogen ion concentration is 0·125 X 0·01 = 1·25 X 10-1 •
Now log 1·25 = 0·097 ;
pH = - ( - 3 + 0'097) = 2·903.
The hydroxyl ion concentration may be expressed m a
similar way:
POH = -loglo [OH-] = loglo i/[OH-1, or [OH-] = 10- OR.
If we write the equation:
[H+] X {OR'1 = Kw = 10-14 ,
in the form:
log [H+] + log [OH-1 = log Kw = -14,
then pH + pOH = pK", = 14.
This relationship should hold for all dilute solutions at about
25°C.
Fig. 1-3 will serve.as a useful mnemonic for. the relation be-
tween [H+], pH, [OH-] and pOH in acid and alkaline solution.
pH I
pOH ~ ~ ~ " W 9 876643210
[OH-]10'4 1()'3 10'2 lOU 1610 109 10
8
16
7
10
8
lOll 164 3
16 162 16' 1 (1d')
--------A-C-id-------+t~.-------AI-ka-li-n.------
Neutral
FIG. 1-3.
The logarithmic or exponential method has also been fou¥>-

useful for expressing other small quantities which arise in
quantitative analysis. These include (i) dissociation con-'
stants (Section I, 4), (ii) other ionic concentrations, and (iii)
solubility products (Section I, 8).
(i) For any acid with a dissociation constant of K,. :
pK,. = - log K,. = IJlog Ka.
Similarly for any base with dissociation constant Kb :
PKb = - log Kb = 1jlog K b.
(ii) For any ion I of concentration [Il :
pI = - log [I] = Ijlog [I].,
J:hus for [Na+] = 8 X 10-5, pI _= 4·1.
(iii) For a salt with a solubility product 5 :
ps = - log 5 = 1 Jlog S.
I, 18. The hydrolysis of salts.-Salts may be divided into
four main groups:
I. those derived from the strong acids and strong bases,
e.g., potassium chloride; ,
II, those derived from weak acids and strong bases, e.g.,
sodium acetate;
III, those derived from strong acids and weak bases, e.g.,
ammonium chloride; and
IV, those derived from weak acids and weak bases, e.g.,
ammonium formate or aluminium acetate.
When any of these is dissolved in water, the solution, as is
well known, is not always neutral in reaction. Interaction
occurs with the ions of water and the resulting solution may be
neutral, acid or alkaline according to the nature of the salt.
With an aqueous solution of a salt of group I, neither the
anions have any tend'ency to combine with the hydrogen
ions nor the cations with the hydroxyl ions of water since the
rel~ted acids and bases are strong electrolytes. The equili-
brium between the hydrogen and hydroxyl ions in water:
H 2 0 ~ H+ + OH- (i),
is therefore not disturbed and the solution remains neutral.
Consider, however, a salt MA derived from a weak acid
HA and a strong base BOH (group II). The salt is completely
dissociated in aqueous solution:
MA ~M+ +A-.
A very small concentration of hydrogen and hydroxyl ions,
originating ftom the small but finite ionisation of water, will
be initially present. HA is a weak acid, that is, it is disso-
ciated only to a small degree; the conceI}tration of A-ions
which can exist in equilibrium with H+ ions is accordingly
small. In order to maintain the equilibrium, the large
initial concentration of A- -ions must be reduced by combina-
tion with H+ ions to form undissociated HA:
H+ + A- ~ HA (ii) .
The hydrogen ions required for this reaction can be obtained
only from the further dissociation of the water; this dissocia-
tion produces simultaneously ah equivalent quantity of
hydroxyl ions. The hydrogen ions are utilised in. the forma-
tion of HA, consequently the hydroxyl ion concentration of
the solution will increase and the solution will react alkaline.
The net result is that the anions of the salt react with t~e
hydrogen ions of the water yielding the weak acid HA, and
there is an increase in the concentration of hydroxyl ions over
that present in water.
32 Quan}itative Inorganic Analysis
It is usual in writing equations involving equilibria ,between
completely dissociated and slightly dissociated or sparing1y
soluble substances to employ the ions of the former and the
molecules .of the latter. The reaction is therefore written:
(iii) .
This equation can also be obtained by combining (i) and (ii)
since both equilibria must coexist.
This interaction between the ion (or ions) of a salt and the
ions of water is called hydro~sis. Formerly the chemical.
reaction was written:
MA + H 20 ~ MOH + HA,
or as:. Salt + Water ~ Base + Acid.
The reaction of the solution was clearly dependent upon the
relative strengths of MOH and HA. This led to th~ definititm
of hydrolysis as the decompdsition of a salt 'into an acid and
a base.
Let us now study the salt of a strong acid and a weak base
(group III). Here the initial high concentration of cations
M+ will be reduced by combination with the hyqroxyl ions of
water to form the littlfT dissociate<;l base MOH until the
equilibrium:
(iv),'
is attained. The hydrogen ion concentration of the solution
will thus be increased and the solution will rea:ct acid. The
hydrolysis is here represented by :
M+ + H 2 0 ~ MOH + H+ (v).
For salts of group IV in which both the acid and the base
are weak, two reactions will occur simultaneously:
M+ + H 20 ~ MOH + H+ (vi),
A - + H 2 0 ~ ItA + OH- (vii).
The reaction of the solution, will clearly depend upon the
relative dissociation constants of the acid and the base. If
they are equal in strength, the solution will be neutral; if
K" > Kb, it. will be acid, and if Kb > K", it will -be alkaline.
Having considered all the possible cases, we are now in a
position to give a more general definition of hydrolysis.
Hydrolysis is the interaction between the ion (or ions)' of a
salt and the ions of water with the production of (a) a weak
acid or a weak base, or (b) of both a weak acid and a weak
base.
The Theoretical Basis of Quan,titative Analysis 33
I, 1~. Hydrolysis constant and degree of hydrolysis.-
Case I. Salt of a weak acid and a strong base. The equili-
brium in a solution of a salt MA may be represented hy :
M+ + A- + H 20 ~ M+ + OH- + HA,
or by A- + HIP ~ OH- + H~ (i).
Applying the law of mass action, we obtain:
aOH- X aHA' [OH-]. [HA] fOH-' fHA
= X = KL (ii)
~ ~] ~ .,
"there a, f, and [ ] refer to activiti::!s, activity coefficients
and concentrations respectively, and K" is the hydrolysis
constant. The solution is ,assumed to be dilute; the activity
of the unionised water may be taken as constant. In dilute
solutions, the ionic strength is small, and the approximation
that the activity coefficient of the unionised acid is unity and
that both ions have the same activity coefficient may be
introduced. Equation (ii) then reauces to :
K _ [OH-] X [HAJ
(iii).
" - [A]
This is often written in the form:
K" = [Base1 X [Acid] (iv) ;
[Unhydrolysed salt]
the free strong base and the unhydrolysed salt are completely
dissociated and the acid is very little dissociated.
The degree of hydrolysis is the fraction of each gram mole-
cule hydrolysed at equilibrium. Let 1 gram 'mol of sals be
dissolved in V Htres of solution and let x be the degree of
hydrolysis. The concentrations in gram mols OF gram ions
per litre are: •
A- + H 2 0 ~ OH- + HA
(I-x )tV • ~/V xW. .
Substituting these values, in (iii) :
•
K _ [OH-] X [HAl _ xjV X xjV _ X2 ( )
II - [A-] - (1 -x)/V - (l-x)V v.
This expression enables us to calculate the degree of hydrolysis

from the value of the )lydrolysis constant at the dilution V.
It is eviden~ that as V increases, the degree of hydrolysis x
must increase.
c
The two equilibria :.
H 2 0 ~ H+ + OR-,
and ~ 'HA ~ H+ + A-,
must coexist with the hydrolytic equilibrium:
A - + H 2 0 ~ HA + OH-.
Hence the two relationships:
[H+] X [OH-] K",
and [H+] X [A -] j [HAl - Ka.
must hold in the same solution as :
[OH-] X [HA] j [A -] = Kil.
K", _ [H+] X [OH-] X [HA] _ [OH-] X [HA] _ K
But Ka - [H+] X [A-] - [A ] - h,
therefore K",jKa =,Kh (vi),

or pK" = pK", - pKa (vii).
The hydrolysis constant is thus related to the ionic product of
the water and the ionisation constant of the acid. Since Ka
varies slightly and K", varies considerably with temperature,
KI> and consequently the degree of hydrolysis will be largely
influenced by changes of temperature,
Example II. Calculate (i) the hydrolysis constant, (ii) the degree
of hydrolysis" and (iii) the hydrogen ion ~oncentration of a 0·01
molar solution of sodium acetate at the laboratory temperature.
K" 1·0 X 10-14
Kb = Ka = 1.82 ><10- 5 = 5·5 X 10-1 °.
The degre£ of hydrolysis x is given by :
x2
K" = (1 -x)V
Substituting for K" and V (= lIe), we' obtain:
x 2 X 0·01
5·5 X 10-10 = .
• (1 -x)
Solving this quadratic equation for x, x = 0·000235 or 0'0235%.
C2 H a0 2- + H 20 ~ H.C2H a0 2 + OH-.
(1 - x) mols x mols x mols
If the solution were completely hydrolysed, the concentration of
acetic acid produced would be O·OIM. But the degree of hydrolysis
is 0·0235%, therefore the concentration of acetic acid is 2';35 X 10-8
The Theoretical Basis oj Quantitative Analysis 35
M. This is also equal to the hydroxyl ion concentntion produced,
i.e., POH = 5·63.
pH = 14·0 - 5·63 = 8·37.
The pH may also be computed from equation (ix) below:
+
pH = 1/2 pKw + 1/2 PK" 1/2 log c,
= 7·0 +
2·37 +
1/2 (- 2) = 8·37.
The hydrogen ion concentration of a solution of a hydrolysed
salt can be readily computed. The amounts of HA and OH-
ions formed as a result of hydrolysis are equal, th~refore in a
solution of the ~ure salt in water [HA] = [OH-]. If the
concentration of the salt is c gram mols per litre, then:
[HA] X [OH-] _ [OH-P _ K _ K."
[A ] - c - ~- K,,'
and [OH-] = yc. Kw I K" (vii),

or [H+] = vK.". ~a / c, since [Ji+] = K", I [OH-] (viii);
PH = If2pKw + 1/2pKa + 1/2logc * (ix).
Equation (ix) can be employed for the cal~ulation of the
pH of a solution of a salt of a weak acid and a strong base.
Thus the pH of a O'05N solution of sodium benzoate is given
by:
pH = 7·0 + 2·10 + 1/2( - 1·30) = 8'45.
(Benzoic acid: Ka:- 6·37 X 10-5 ; pK" = 4·20.)
Such a calculation will provide useful information as to the
indicator whiCh should be employed in the titration of a weak
acid and a strong base (see Section I, 33).
case 2. Salt of a strong acid a.nd a weak base. The hydro-
lytic equilibrium is represented by :
M+ + H 2 0 ~ MOH + H+ (x).
By apply~ng the law of mass action along the lines of Case 1.
the following equations are obt~ined : •
K" = [H+] X [MOI{] = [Acid] X [Base] K",
[M+] [Unhydrolysed Salt] = Kb (xi),
(xii).
(1 - x)V
• :ro be consistent we should write pc = - log c.
Kb is the dissociation constant of the base. Furthermore,
since [MOH) and. [H+) are equal (equation (x),
K _ [H+) X [MOH] _ [H+)2 _ K
"- [M+] - -c- - Kb '
[H:'-} = v'c.K", I K b , •
()r pH = 1/2PKfII _ 1/2 PKb - 1/2 log c (xiii) •
Equation (xiii) may be applied to the caloulation of the
pH of solutions of salt>s of strong acids and weak bases. Thus
the. PH of a O·02N solution of ammonium chlQride is :
pH = 7·0 _ 2·37 _ 1/2 (-0'70) .. 4: 98.
(Ammonium·hydroxide: Kb = 1·85 X 10-0 ; PKb = 4'74.)
. Case 3, Salt of a weak acid and a weak base. The hydro-
lytic equilibrium is expressed by the equation:
M+ + A- + H 0 ~ MOH + HA (xiv) .
. 2
Applying the law of mass action and taking the activity.of

unionised water as u~ity, we have:
K" = aMOH X aHA = [MOH]. [HAl X /MOH • /HA
aM+ X aA- [M+] . [A ] /M+ . f A- (xv).
By the usual approximations, that is, by assuming that the
activity coefficients of the unioni'sed molecules and, less
justifiably, of the ions are unity, the following approximate
equation is obtained:
K _ [MOH1 x {HAl (xvi),
,,- [M+] X [A ]
_ [Unhydrolysed
[Base] X [Acid]
--~--~~--~-
Saltp
(xvir)~
.
If x. is the degree of •hydrolysis of 1 gram mol of the salt
dissolved in V litres of solution, then the individual concen-
trations are:
[MOR] = .JHA] = xiV; [M+] = [A -1 = (I - x)/Y.
Substituting these values in (xvi) :
K,,= _ xIV. xIV _ _ X_2_
(xviii) .
(1 - x)/V.(l - x)/V (1 _" X)2
The degree of hydrolysis and consequently the pH is inde-
pendent of the..concentration of the solution.*
As in Case 1 the expressions :
Kw = [H+] X [OH-],
K - [H+] X [A-] d K _ [M+]X [OH-]
,. - [HA ,.an b - [MOH] ,
hold simultaneously with equation (xiv) for the hydrolytic
equilibrium. By substitution in ·the last-named equation,
it can be readily shown that:
K" = Kw / K,. X Kb (xix),
or pK" = pKw - PK" ...... PKb (xx).
This expression enables us to compute the value of the degree
of hydrolysis x from the dissociation constants of the acid and
the base.
The hydrogen ion concentration of the hydrolysed solution
is calculated in the following manner~: -
+ _ [HAl _ xIV _ x
[H ] - K .. X [A-] - K,. X (I _ x) IV - KtJ X (1 _ x)·
But x/{1 - x) = IYK; (by equation (xviii) ),

hence [H+] = K,.. vK" = vKw X K,. f Kb . (xxi),
or , pH. If2pKw + IJ2pK,.·- I/2PKb (xxii).
If the ionisation constants of the acid and the base are equal,
that is, K,. = K b , pH = 1/2 pKw = 7·0; and the solution is
neutral although hydrolysis may' be. considerable. If
1(. > K b , pH < 7 and the solution is acid, but when
K. > K,. , pH > 7 and the solution reacts alkaline.
The pH of a solution of ammonium acetate is given by :
pH = 7·0 + 2·37 - 2·37 = 7·0,
i.e., the solution is approximately neutral. On the other
hand, for a dilute solution of ammonium formate:
pH = 7·0 + 1·88 - 2·37 = 6·51,
(Formic acid: 1}.. = 1·77 X 10-4; PK.. = 3·75.)
i.e., the solution reacts slightly acid .
•• This applies only 'f the original assumptions as to activity coefficients are
justified. In solutions crf appreciable ionic strength, the activity coefficients
of the ions will vary with the total ionic strength.
38 Quantitative lnorgimic Analysis
I, 20. Buffer solutions.-A solution of 0, oo()1N hydro-
chloric acid should have a pH equal to 4, but the solution is
extremely sensitive to traces of alkali from the glass of the
containing vessel and to ammonia from the air. Likewise a
O·OOOIN solution of sodium hydroxide, which should have a
pH of 10, is sensitive to traces of carbon dioxide from the
atmosphere. Aqueous solutions of potassium chloride and
of ammopium acetate have a pH of about 7. The addition of
I m!. of normal hydrochloric acid to I litre of the solution
results in a change of pHOto 3 in the former case and in very
little change in the latter. The resistance ,of a solution to
changes in hydrogen ion concentration upon the addition of,
small amounts of acid or alkali is termed buffer action; a
solutipn which possesses such properties is known as a buHer
solution. It is said to possess" reserve acidity" and" reserve
alkalinity." Solutions of which the PH values (determined
by reference to the hydrogen electrode) are known, which can
be readily prepared, and which are unaffected by small
additions of alkali or acid, are required for the colorimetric
determination of hydrogen ion concentration (see Section
V, 6) and for other purposes.
Buffer solutions usually consist of solutions containing a
mixt\lre of a weak acid or base and its salt. In' order to
understand buffer action, let us study first the equilibrium
between a weak acid and its salt. The dissociation of a weak
acid is given by : '
HA ~ H+ + A-,
and its magnitUde is controlled by the value of the dissociation
constant Ka :
(i) ,
where a" refers to the activity of the species x. The expres-

sion may be approximated by writing concentrations for
activities (strictly speaking, it will be recalled (Section I, 5),
activity = co;ncentration X activity coefficient) :
[ H+J = [HAJ K (ii) .

[A -] X a
This equilibrium applies to a mixture of an acid HA arid its

salt, say, MA. If the concentratiori of the acid be Ca and that
of the salt be c., then the conc,entration of the undissociated
portion of the acid is Ca - [H+]. The solution is electrically
neutral, hence [A-] = c. + [H+] (the salt is completely dis-
sociated). Substituting these values in the equilibrium
equation (ii), we have:
c - [H+]
[H+] = a X K (iii).
c. + [H+] a
This is a quadratic equation for [H+] and may be solved in the

usual manner. It can, however, he simplified by introducing
the following further approximations. In a mixture of a
weak acid and its salt, the dissociation of the acid is repressed
by the common ion effect, and [H+] may be taken as ~egli
gibly small by comparison with Ca and c.. Equation (iii)
then' reduces to :
[ H+
.]
= C~ ' K
a, or
[H+] = [Acid] K
[Salt) X a
(iv),
s
[Salt]
or pH = pKa + log [Acid) (v).
Similarly for a mixture of a weak base of dissociation
constant Kb and its salt with a strong acid:
[ OH-] = [Base] X K (vi)
[Salt] b
or POH = PKb + log [~::?] (vii).

Let us confine our attention to the case in which the con-
centrations of the acid and its salt are equal, i.e., of a half
neutralised acid. Then pH = pKa . Thus the pH of a_
half neutralised solution of a weak acid is equal to the nega-
tive logarithm of the dissociation constant of the acid. For
acetic acid, Ka = 1·82 X 10-5 , PKa = 4'74; a half neu-
tralised solution of, say, O'lN acetic acid will have a pH of
4·74. If we add a small conc~ntration of H+ ions to such a
solution, the former will combine with the acetate ions to
form undissociated acetic acid: . ':to
H+ + C2H 3 0 2- ~ H.C 2H a0 2 •
Similarly, if a small concentration of hydroxyl iQns be added,
the.. latter will combine with the hydrogen ions arising from
the dissociation of the acetic acid and form unionised water;
the equilijJrium will be disturbed. and more acetic acid will
dissociate to replace the hydrogen ions removed in th}s' way.
In either case, the' concentration of the acetic acid and acetate
ion (or salt) 'will not be appreciably changed. It follows
from equation (v) that the pH of the solution will not be
materially affected.
The solution containing equal concentrations of acid and
its salt, or a half neutralised solution of the acid, has the
maximum buffer capacity. Other mixtures also possess
considerable buffer capacity, but the pH will differ slightly
from the half neutralised acid. Thus in a.quarter neutralised
solution of acid, [Acid] =.3 [Salt]: -
pH = PK" + log 1/3 = pK" + 1'52,
=pK" - 0·48.
For a three-quarter neutralised acid, [Salt] = 3 [Acid] :
pH = pK" + log 3,
=pK" + 0·48.
In general, we may 'state that the: buffering capa~ity is main-
tained for mixtures within the range I acid: 10 salt and 10
acid: .1 salt. The approximate pH range of a weak acid
buffer is: •
pH = pK" ± 1.
The conGentration of the acid is usually of the order 0·05 -;- 0·2
molar. Similar remarks apply to weak bases. ..
The preparation of a buffer solution of a definite PH is a
simple process if the acid (or base) of appropriate dissociation
constant is found: small variations in pH are obtained by
variations in the ratio of the aci$l to the salt concentration.
One example is given in Table VII,
TABLE vn. pH OF ACETIC ACID-SODIUM ACETATE BUFFER MIXTURES.-
l~ ml. mixtures of x ml. of O·2M acetic acid and y m!. of O·2M sodium acetate .
Acetic Acid (x ml.). .$odi~m Acetate (y ml.) .PH

9·5 0'0 3·42
9·0
-g·0
1·0
2·0
3·72
4·05
.
7·0 3·0 4·27
6·0 4·0 4·45
5·0 5·0 4·63
4·0 6·0 4·80 .
3·0 ,7·0 4'99
2·0 8·0 5·23
1·0 9·0 5·57
• 0·5 9·5 5·89
-. See ..(\ppendix, Section A, 10.
The Theoretical Basis of Quantitative A.nalysis 41
BeJore leaving the subject of buffer solutions, it is neces-
sary to draw attention to e. pos&ible ~rroneous deduction from
equation ~v), 'namely, that the hydrogeh ion concentration
of a buffer solution is depeQlient only 'upon the ratio of the
concentrations of acid.and salt and upon K a , and not upon
the actual concentration~; otherwise expressed, that the
pH of such a buffer mixture should not change upon dilution
with water. This is approximately although not strictly
true. In deducing equation (ii), concentrations have been
substituted for activities\ a step which is not entirely justi-
fiable except in dilute solutions. Theoretically,.the expression
controlling buffer action is :
aHA c... f" •
aH+ = - X K" = - _ X K" (viii)
aA- Cs.fA-'·
where a and f refer to activities and act,ivity coefficients
respectively of the species indicated in the subscript. The
,activity coefficient: fa of the undissociated -acid is approxi-
mately unity in dilute aqueous solu~ion. Expression (viii)
thus becomes :
[Acid] X K (ix),
aH+ = [Salt] X fA- a
or pH = PK.. +"log [Salt]/[Acid] + log !A- (x).

According to the Debye-Hiickel-Onsager theory, the activity
coefficirnt ot an ionfi in aqueous solution at 25°C is given by :
0'505z,21L 0·&
log/. = -
1 + .
A1L 0 • 5
+ BIL,
where Z; is the valency of the ion, IL is the ionic strength of the

solution, and A and B are constants. This may be written
in the form:
log!; = - 0·505 Z;2 IL O.5 + CIL,
where C is another constant approximately equal to 0·505 z;2A
+ Band usu<,\lly has a value varying between 0'2 and
1'5. Substituting for J~- in (v), we obtain:
pH = PKa + log [Salt] I [Acid] - 0·505z;21L 0.5 + CIL (xi).
The activity coefficient of the ion fA- generally iijcreases
with decrease of concentration, so that when a buffer solution
is diluted, fA- increase§ and consequently aH+ will increase
(equation (ix) ) .. For most pr!lctical purposes the change is
small, but for exact work the change must be taken into
account. The addition of salts to buffer m.ixtures results in
c*
a change of the ionic strength of the solution; this will affect
the pH of the solution (~quation (xi)). Indeed, in all buffer
solutions a correction should, strictly speaking; be applied for
the ionic strength of 'the solution.
Buffer mixtures are not ctmfined to mixtures of monobasic
acids or monoacid base:; and their salts. We may employ a
mixture of salts oL a polybasic acia, e.g., NaH 2P0 4 and
Na 2 HP0 4 • The salt NaH 2 P0 4 is completely dissociated:
NaH 2 P0 4 .= Na+ + H 2 P0 4-.
The ion H 2 PO 4- acts as a monobasic acid:
H 2 P0 4- .= H+ + HP.0 4--,
for'"Which K (sK 2 for phosphoric aeid) is 6·2 X 10- 8 • The
addition of the salt Na 2 HP0 4 is analogous to the addition of,
say, acetate ions to a solution of acetic acid, since the tertiary
ionisation of phosphoric acid (HP0 4- - .= H+ + P0 4- - - )
is small (K3 = 5 X 10-13). The .mixture of NaH 2 P0 4 and
Na 2 HP0 4 is Ltherefore an effective buffer over the range pH
7·2 ± 1'0 (:____ PK±l). .
Buffer solutions find mahy applications in quantitative
analysis. Many precipitations are made in certain ranges of
pH values. Examples are given in Sec1ion I, 61 under
Separations; many others will b~ found scattereq. throughout
the text. Buffer solutions are ~lso employed in the colori-
metric determination:of pH with the aid of indicators CSection
V,6).
VOLUMETRIC ANALYSIS
In the foregoing pages we have studied some general theory,
a knqwledge of which is indispensable for the comprehension
of the processes of quantitative analysis, both gravimetric
and volumetric. We will now confine our attention to the
principles of vQlum~tric analysis and postpone the parallel
study of gravimetric analysis to a later stage (Section 1,54).
I, 21. Volumetric analysis, or quantitative chemical
analysis by measure, consists essentially in determining the
volume of a solution of accurately known cQncentratism which
is required to react quantitatively with the solution of the
substance being determined. The solution of accurately
known 1;_trength is called the standard Bolution; it contains a
definite ·number of gram equivalents (see Section I, 23) per
litre. The weight qf the substance to be determined is then
calculated from the volume of the standard solution and the
known laws of chemical equivalence.
The standard solution is usually added from a graduated
vessel, called a burette. The process of. adding the standard
solution until the reaction is just comp~ete is tfirmed a
titration, and the substance to be determined is titrated. The
point at which this occurs is called the equivalence point or the
theoretical (or stoichiometric) end pOint. The completion of
the titration should, as a rule, be detectable by some change
unmistakC1ble to the eye produced by tlie standard solution
its~lf (e.g., potassium permanganate) or, more usually,_by the
adaition of an auxiliary reagent, known as an indicator.
After the reaction between the substance and the standard
solution is practically complete, the indicator should give a
clear visual change (either a colour change or, the forlV-ation
of a turbidity) in the liquid being titrated. The pomt' at
which this occurs is called the end point of the titration. In
the ideal titration the visible end point will coincide with the
stoichiometric or theoretical end point. In practice, how-
ever, a very small difference usually occurs; this represents
the titration error. We should always endeavour to select an
indic.ator and also the experimental conditions such that the
difference between the visible end point and the equiV'alence
point is as small as possible.
For use in volumetric analysis a reaction must fulfil the
following conditions:
43
1. There must be a simple reaction which can be expressed
by a chemical equation; the substance to be determined
should react completely with the reagent in stoichiometric
or equivalent proportions.
2. The reaction should be practically instantaneous or
proceed with very great speed. "(Most ionic reactions satisfy
this condition.) In some cases the additiQn of a catalyst
increases the speed of a reaction.
3. There must be a marked change .in some physical or
chemical property of the solution at the equivalence point.
4. An indicator should be available which, by a change in
physical properties (colour or formation of a precipitate),
should sharply define the end point of the reaction. [If no
visible indicator. is available for the'detection of the equiva-
lence point, the latter can often be determined by fpllowing
during the course of the titration (a) the change of the
potential of a suitable electrode (potentiometric titration, see
Sections I, 53), or (b) the change of the electrical conductance
of the solution (conductometric titration).]
Volumetric methods are, as a rule, susceptible of high
accuracy (1 part in 1000) and possess several advantages,
wherever applicable, over gravimetric methods. They need
simpler apparatus, tedious and difficult separations can often
be avoided, and are, generally, quickly performed. The
following are required for volumetric analysis: (i) Calibrated
measuring vessels including burettes, pipettes and measuring
flasks (see Sections n, 20-22). (ii) Substances of known purity
for the pJ;eparation of standard solutions. (iii) An indicator
or other device (see Section I, 53) for detecting the complete-
ness of the reaction.
I, 22. Classification of reactions in volumetric analysis.-
The reactions employed in volumetric analysis fall into two
m:Hn Classes: •
(a) Those in which no change in valency occurs; these are
dependent upon the combination of ions.
(b) Oxidation-reduction reactions; these involve a change
of valency'or, otherwise expressed, a transfer of electrons ..
For purposes of convenience, however, these two types gf
reactions are divided into three main classes :'
1. Neutralisation reactions.- or acidimetry and alkalimetry.
These include the titration of free bases or those formed from
salts of weak acids by hydrolysis with a standard acid (acidi-
metry), and the titration of free acids or those formed by the
The Theoretical Basis of 9uantitative. Analysis 45
hydrolysis of salts of weak bases with a standard base (alkali-
metry). These reactions involve the combinat'ion of hydrogen.
and hydroxyl ions to form water. • -
2. Precipitation reactions. These depend upon the com-
bination of ions, other than hydrogen and hydroxyl ions, to
form either. a simple precipitate, as in the titration of silver
with a solution oJ a chloride (Section m.
24). or a complex
ion, as in the titration of a cyanide with silver nitrate ~olution
(Section m. 41; 2CN- + Ag+ ~ [Ag(CN)zr).
3. Oxidation-reduction reactions. Under this heading are
included all reactions involving change in oxidation number
(Section I. 23) or transfer of electrons (Section I. 23) among the
reacting substances. The standard solutions are either
oxidising or reducing agents. The principal oxidising agents
are potassium permanganate, potassium dichromate, ceric
.sulphate, manganic sulphate, iodine, potassium iodate, potas-
sium bromate and chloramine-To Frequently used reducing
agents aret'ferrous and stannous compounds, sodium thiosul-
phate, arsenious oxide and titanous chloride or sulphate.
I
1.23. Equivalent weights. Normal solutions.-A-standard

solution is one-which contains a known weight of the reagent
in' a definite volume of the solution. ·A molar solution is one
which contains'one gram molecular weight of the reagent per
litre of solution. A normal solution of a reagent is one that
contains one gram. equivalent weight* per litre of solution.
It is designated by N. Semi-normal, .penti-normal, deci-
nprmal and centi-normal solutions are often required; these
are shortly written·N/2 or 0·5N, N (5 or, 0·2N, N /10 or O'lN,
and N {100 or O·OlN respectively. .
The above definition of normal solution utilises the term
" equivalent weight." This quantity varies with the type of
reaction, and, since it is difficult to give a clear definition of
"equivalent weight" which will cover all reactions, it is
proposed to discuss this subject in some detail below. It often
happens that the same compound: possesses different equiva-
lent weights in different chemical reactions. The situation
may therefore arise in which a solution may have normal
concentration _when employed for one purpose, an,d a different
normal concentration when used in another chemical reaction.
Neutralisation reactions.-The equivalent weight of an acid
is that weight of it which contains one replaceable hydrogen,
* Sometimes also called equivalent, gram equivalent or equivalent weight.
46 Q,uantitative Chemical Analyst's
i.e., 1'008 (more, accurately 1·0078) grams of liydrog~n. The
equivalent weight of a monobasic acid, such as hydrochloric,
hydrobromic, hydriodic, nitric, perchloric or acetic acid, is
identical with its molecular weight. A normal solution of a
monobasic, acid will therefore contain one gram molecular
weight (or 1 mol) in a litre o~ solution. The equivalent weight
of a dibasic acid (e.g., sulphuric or oxalic acid), or'of a tribasic
acid (e.g., phosphoric acid) is likewise 1/2 and 1/3 respectively
of its molecular weight.
The equivalent weight of a base is that weight Qf it which
contains one replaceable hydi-oxyl.group, i.e., 17 ·008 grams of
ionisable hydroxyl. 17 ·008 grams of hydroxyl are equivalent
to 1·008 grams of hydrogen. The equivalent weights of
sodium nydroxide, potassium hydroxide and ammonium
hydroxide are 1 mol, of calcium hydroxide, strontium hydroxide
and barium hydroxide 1/2 mol.
Salts of strong bases and weak acids posses's alkaline
reactions in aqueous solution because of hydrolysis (Section
I, 18). Sodium carbonate, with methyl orange as indicator,
reacts with two mols of hydrocploric acid to form tW(J mols of
sodium. chloride; hence its equivalent )Veight is 1/2 mol.
Borax, under similar conditions, also reacts with two mols of
hydrochloric acid, and its equivalent weight is, likewise, 1/2
mol.
Precipitation reactions.-Here the equivalent weight is the
weight of the substance which contains or reacts with one
gFam atom cif a univalent metal (which is equivalent to 1·008
grams of hydrogen), 1{2 gram atom of a bivalent metal, eJc.
For a mf!tal, the equivalent weight is the atomic weight
divided by the valency. For a reagent which reacts with this
metal, the equivalent weight is the weight of it which reacts
with one equivalent of the metal. The reactive constituents
of a salt are its ions. The equivalent weight of a salt in a
precipitation reaction is the gram molecular weight of the
salt divided by the total valency of the reacting ion. Thus
the equivalent weight of silver nitrate in the titration of
chloride is its molecular weight.
In a complex formation reaction, the equivalent weight is
most simply deduced by writing down the ionic equation of
the reaction. Fo~ example, the equivalent weight of potas-
sium cyanide in the titration with silver ions is two mols,
since the reaction is :
2CN- + Ag+ ~ [Ag(CN)2r'
The Theoretical Basis qf Quantitative Analys~s 47
Oxidation-reduction reactions.-The equivalenf weight of an
oxidising or reducing agent is most simply defined as that
weight of the reagent which reacts with or contains 1·008
grams of available hydrogen or 8·000 grams of available
oxygen. By" available" is meant capable of being utilised
in oxidation or reduction. The amount of available oxygen
may be indicated by writing the hypothetical equation, e.g.,
2KMn0 4 = K 2 0 + 2MnO + 50,
i.e., in acid solution 2KMn0 4 gives up 5 atoms of availabfe
oxygen which is taken up by the reducing agent, hence its
equivalent weight is 2KMn0 4 /10. For potassium dichromate
in acid solution, the hypothetical equation is :
K 2Cr 20 7 = K 20 + Cr2 0 3 + 30.
The equivalent weight is K 2Cr 20 7 /6. This elementary treat-
ment is limited in application, but is useful for beginners.
A more general view is obtained by a consideration of (a)
the number of electrons involved in the partial ionic equation
replesenting the reaction, and (b) the change in the" oxida-
tion number" of a significent element in the oxidant or
reductant. Both methods will be considered in some detail.
In quantitative analysis we are chiefly concerned with
reactions which take place in so1ution, i.e., ionic reactions. We
shall therefore limit our discussion of oxidation-reduction to
su({h reactions. The oxidation of ferrous chloride by chlorine
in aqueous solution may be written:
2FeCl 2 + C1 2 = 2FeC1 3 ,
or may be expressed ionically :
2Fe++ + C1 2 = 2Fe+++ + 2Cl-.
The ferrous ion Fe++ is converted into the ferric ion Fe+++
(oxidation), and the neutral chlorine molecule into negatively
charged chloride ions CC (reduction). According to the elec-
tronic conception of the" constitution of matter, the conversion
of Fe++ into Fe+++ requires the loss of one electron, and the
transformation of the neutrOal chlorine molecule into chloride
ions necessitates the gain of two electrons. This leads to
the view that, for reactions in solution, oxidation is a process
involving a loss of electrons, as in
Fe++ - E = Fe+++,
and reduction is the process resulting in a ga;n of eledrons,
as in
In the actual oxidation-reduction process electrons are
transferred from the reducing agent to the oxidising agent.
This leads to the following definitions. Oxidation is the
process which results in the loss of one or more electrons by
atoms or ions. Reduction is the process which results in the
gain of one or more electronsvby atoms or ions. An oxidising
agent is one t~at -gains electrons and is reduced to a lower
valency condition. A reducing agent is ,on~ that lo,ses elec-
trons and is oxidis~d to a higher valency condition.
The following are examples of oxidising agents which are
of importance in quantitative analysis: potassium perman-
ganate, potassium dichromate, ceric sulphate, manganic sul'-
phate, iodine, potassium bromate, potassium iodate and chlora-
mine-T. Examples of reducing agents are: ferrous sulphate,
metallic iron, sodium j:hiosulphate, sodium arsenite or arseni-
ous oxide, oxalic acid, titanous chloride or sulphate and
cliromous sulphate. '
In all oxidation-reduction processes (or redox 'Processes)
there will be a reactant undergoing oxidation and one
undergoing reduction, since the two reactions are comple-
mentary to one another and occur simultaneously-one cannot
take pl~ce without the other. The reagent-suffering oxidation
is termed tlle reducing agent or reductant, and the reagent
undergoing reduction is called the oxidising agent or oxidant.
The study of the electron changes in the oxidant and reductant
forms the basis of the ion-electron method for balancing ionic
equations, a method developed in some detail by Jette and
La ¥er in 1921. The equation is accordingly first divided
into two balanced, partial equation:> representing the oxida-
tion and reduction respectively. It must be rememb&ed
that the reactions take place in aqueous·solution so tnat in
addition to the ions supplied by the oxidant and reductant,
the molecule of water H 20, hydrogen ions H+ and hydro~l
ions OH- are also present, and may be utilised in balancing
the partial ionic equation. The unit change in oxidation. or
reduction is a charge of one electron, which will be denoted by
E. To appreciate the principles involved, let us consider first
the reaction between ferric chloride and stannous chloride in
aqueous solution. The partial ionic equation for the reduc-
tion is :
Fe+++ _,.. Fe++ (a),
'and for the oxidation is :
Sn++ _,.. Sn++++ (b).
The Theoretical Basis. of Quantitative Analysis 49
The equations must be balanced not only with regard to the
number ,and kind ·of atoms, but also electrically, that is, the
net electric charge on each side must be the same. Equation
(a) can be balanced by adding one electron to the left hand
side: •
Fe+++ + E ~ Fe++ (a'),
and equation, (b) by adding two electrons to the right hand
side:
Sn++ ~ Sn++++ + 2 E (b').
These partial et}uations must then be .multiplied by' coeffi-
cients which result in the number of electrons utilised in one
reaction being equal to those liberaled in the other. 'Thus
equation (a ') must be multiplied by two, and we have:
2Fe+++ + 2E ~ 2Fe++ (a") ;
Sn++ ~ Sn++++ + 2 E (b ").
Adding (a") and (b"), we obtain: .,
2Fe+++ + Sn++ + 2 E ~ 2Fe++ + Sn++++ + 2 E,
and by cancelliJ}.g the electrolls common to both sides, the
simple ionic equation is ob.tained:
2Fe+++ +.Sn++ :::= 2Fe++ + Sn++++.
•
The following facts must be borne" in mind. All strong
electrolytes are completely dissociated; hence only the ions
actually taking part or resulting from the reaction need
appear in tJle equation. Substances whifh are only slightly
ionised, ;,uch as water, or which are sparingly soluble and thus
yield only a small concentration -of ions, e.g., silver chloride
and barium sulphate; are, in general, written in tl1e molecular
form because'most of th,e substance is present in this state.
The complete rules for the application of the ion-electron
method may be expressed as follows':-- .,
(1) Ascertain the products of the reaction .
. (2) Set up a partial equation for the oxidising agent.
(3) Set up a partial equation for the redu~ing agent in the
same way. •
(4) MUltiply each partial equation by a factor so that when
the two' are added'the electrons just compensate each other.
(5) Add the partial equations and cancel out substances
which appear on both sides of the equation.
A few examples follow.
Reaction I:. the reduction of potassium permanganate in
the presence of dilu.te sulphuric acid. -
50 Quantitative Inorganic Analysis>
The first partial equation (r:eduction) is :
Mn0 4- -+ Mn++.
To balance atomically, 8H+ is required:
#Mn0 4- + SH+ -+ Mn++ + 4H zO;
and to balance it electrically 5 E is needed on the left hand side:
Mn0 4- + SH+ + 5E ~ Mn++ + 4H 20.
The second partial equation (oxidation) is:
Fe++ -+ Fe+++.
To balance this electrically one electron must'be. added to the
right hand side or subtracted from the Ieit hand side-:
Fe++ ~ Fe+++ + E.
Now the gain and loss of electrons must be equal. One
permanganate ion utilises 5 electrons, and one ferrous. ion
liberates I electron; hence the two partial equations must
take place in the ratio of I : 5.
MnO,- + sH+ -f- 5E ~ Mn++ + 4H 2 0
5(Fe++ ~ Fe+++. + E)
or Mn0 4 + 8H+ + 5Fe++ = Mn++ + 5Fe+++ + 4H 20.
Reaction II: the interaction of potassium-dichromate and
potassium iodide in the presence of dilute sulphuric acid.
tr 0 -- ~ Cr+++ ...
Cr 20 7- - + 14H+ ~ 2Cr+++ + 7H 20.
2 7 "
To balance electrically, add 6E to the left hand side:

Cr 2 0 7- - + 14H+ + 6E ~ 2Cr+++ + 7H 2 0.
1:.he various stages in the' deduction of the second partial
equation are:
1- -+ 12 ;
21- -+ 12 ;
21- ~ 1~ + 2E.
One dichromate ion uses 6 E, and two iodide jons liberate 2 E ;
hence the two partial equations take place in the ratio of 1 : 3.
Cr 2 0 7- - + 14H+ + 6'E ~ 2Cr+++ + 7H 2 0
3(2r ~ 12 + 2E)
or Cr 2 0 7 + 14H+ + 61 = 2Cr+++ + 7H 20 + 31 2 ,
We can now apply our knowledge of partial ionic equations
to the subject of equivalents. The standard oxidation-
reduction process is H ~ H+ + E, where E represents an
electron per atom or one Faraday (96,500 coulombs) per grim
atom. If we know. the change in the number 9f electrons I
per ion in any oxidation-reduction reaction, the equivalent

may be cakqlated. The equivalent weight of an oxidant or
reductant is the molecular weight divided by the number of
electrons which one molecule of the substance gains or loses
in the reaction.
We will write down the part~al equations for four fypical
reactions:
Mno.- + 8Jt+ + 5~ = Mn++ + 4H 20. Eq. wt. = MnO,- /5 = KMnO,/5.
Cr20 1- - + 14H+ + 6£ = 2Cr+++ + 7H.O. Eq. wt. = Cr.Q;. --/6 = K.Crp7/6.
Fe++ - £ = Fe+++ Eq. wt.'= Fe++ /1 = FeSO,/l.
c.o,-- - 2£ = 2CO.. Eq. wt. = C.O, --/2 = H.C.O,,I.2.
For convenience of reference the partial ionic equation for a
number of oxidising and reducing agents are collected in
Table VIII. ..
The other procedure which is of value in the calculation of
the equivalent weights of substance is the "oxidant number ?'
method. The student ~hould be thoroughly conversant
with the significance of " oxidation number" and its use in
balancing oxidation-reduction equations.* The equivalent
weight of an oxidising agent is determined by the cliange in
,oxidation number which the reduced element 'experiences.
It is that quantity of oxidant which involVes a change of one
unit in the oxidation-number. Thus in the normal reduction
.of potassium permanganate in the presence of dilute sul-
phuric acid to a manganous salt :
+1+7-8 +2+6-8
K Mn 0, -+ Mn S 0,
the change in the oxidation number of the manganese IS
from +7 to +2.
The. equivalent weight of potassium permanganate is
therefore 1/5 mol or by 5 units of reduction. Similarly for
the reduction of potassium dichromate in acid soh,ltion : .
+2+12-14 +6-6
Crz 0 7 --'>- Cr2 (S04)3
K2
the change in oxidatiOh number oj two atoms of chromi,um is

from + 12 to +6, or by 6 units of reduction. The equivalent
weight of potassium dichromate is accordingly 1/6 mol. In
order to find-the equivalent weight of an oxidising agent, we
divide the molecular weight by the change in oxidation num-
ber per molecule which sOJpe key element in the substance
undergoes.
* A detailed account will be found in the author's Text Book of Qualitative
Chemical Analysis, 1937, p. 42 et seq. (Longmans, Green and Co.) •
52 Quantitative Inorganic Analysis.
TABLE VIII. IONIC EQUATIONS FOR USE IN THE CALCULATION OF THE
EQUIVALENT WEIGHt§ OF OXIDISING AND REDUCING AGENTS.
Substance. Partial Ionic Equation.

OXIDANTS
Potassium permanganate + 8H+ + 5£ ,.= Mn++ + 4H.0
(acid)
Potassium permanganate + 2HsO + 3 £ .= MnO. + 40H-
(neutral)
Potassium permanganate + €.= MnO,!.-
(stronglya:lkaline) I
Ceric sulphate Ce++++ + €.= Ce+++
Potassium dichromate Cr.O.-- + 14H+ + 6 € .= 2Cr+++ + 7H.0
Chlorine Cl. + 2€ .= 2Cl-
Bromine Br. + 2€ .= 2Br-
IodiIte I. • +2E.=21-·
Ferric chloride Fe+++ + €.= Fe++
Potassium. bromatE!" BrO,-' +~+ +' 6€ .= Br- + 3Ha0
Potassium iodate 10a- + 6H+ + 6 € .= r + 3H,0
Sodium hyp'ochlorite CIO- + H.O + 2€ .=cr + 20H-
Hydrogen peroxide H 20, + 2H+ + 2€ .=2H,0
Manganese dioxide MnO. + 4H+ + 2€ .= Mn++ + 2HsO
Sodium bismuthate BiO s- + 6H+ + 2" .= Bi+++ + 3H.0
Nitric acid (conc.) NO a- + 2H+ + ".= NO. + HIO
Nitric a!:id (dilute) NO a- + 4H+ + 3€.= NO 4- 2H.0
REDUCTANT5
.
Hydrogen H. -2E.=2Ht
Zinc Zn - 2£ .= Zn++
Hydrogen sulphide H.S - 2£ .=2H++ 5
Hydrogen iodide 2HI - 2€.= I, + 2H+
Oxalic acid c,o,-- - 2" .=2CO.
Ferrous sulphate Fe++ - ".= Fe+++
SulphurouS'acid HISOl. - 2".= 50,-- + 4H+
Sodium thiosulphate 25,0. - - 2€.= S~O.--
Titanous sulphate Ti+++ ;_ ".= Ti++++
Stannous chloride ~n++ - 2E .= 5n++++
Stannous chloride (in pres-
ence of hydrochloric
acid) _ + 6Cl- - 2£ .= SnCl.--
Hydrogen peroxide - 2£ .=2H+ + 0.
The equivalent weight of a reducing agent is similarly deter-
mined by the change in oxidation number which the- oxidised
element suffers. Consider the conversion of ferrous into
ferric sulphate:
+2 -2
2(FeSO,) -+
Here the change in oxidation number per atom of iron is from

+2 to +3, or -by 1 unit of oxidation, hence the equivalent
weight of ferrous sulphate is I mol. Another important
reaction is the oxidation of oxalic acid to carbon dioxide and
water:
-2 +6-8 +4 -4
H2 C2 0 4 -+ 2C O2
The change in oxidation number of two atoms. of carbon is
from +6 to +8, or by 2 units of oxidation. The equivalent
weight of oxalic acid is therefore 1/2 mol.
In ge~eral, it may be stated:
~i) The equivalent weight of an 'element taking part in an
oxidation-reduction {redox) reaction is the atomic weight
divided by the change in oxidati.on number. •
(ii) When an atom in any complex molecule suffers a
change in oxidation number (oxidation or reduction); the
equivalent weight of the substance is the molecular weight
divided by the change 'in oxidation number of the oxidised
or reduced element. If more than one atom of the reactive
element is present, the,molecular weight is divided by the
total change in oxidation number.
A useful summary of the common oxidising and reducing
agents, together with the various transformations which they
.llndergo, is given in Table IX.
We are now in a position to understand more clearly why
the equivalent ,weights of some substances ~ry with the
reaction. We will consider'two familiar examples by way of
illustration. A normal solution of ferrous sulphate FeS04,
7H 20 will have an equivalent weight of 1 mol when employed
as a reductant, and 1/2 mol when employed as a precipitant.
A solution of ferrous sulphate which is norma] as a precipitant
will be half normal as a reductant. Potassium tetroxalate
KHC204,H2C204,2H20 contains three replaceable hydrogen
atoms.; its equivalent weight in neutralisation reactions is
therefore 1)3 mol:
KHC204,H2C204,2H20 + 3KOH = 2K 2 C2 0" + 5H20.
A~ a reducing agent, a gram molecular weight contai_p.s
2C 2 0 4--, and the equiv?-lent w.eigh~ is accordingly 1/4 mol:
C20 4- - - 2€ . 2C02 •
A solution of the salt which is 3N as an acid is 4N as a reduc-
ing agent.
I. 24. Advantages of the use of the equivalent system for
the preparation of standard solutions.--The most important
advantage of the equivalent system is that the calculations
of volumetric analysis are rendered very simple, since a~ the
54 Quantitative Inorganic A'nalysis
TABLE IX
COMMON OXIDISING AGENTS
Radical O.N. of De-

ar ele- " effec- Reduc- New crease Gain in
Substance tive .. tion O.N. in in elec-
ment trans
involved element product O.N.
KMnO, (acid) MnO,- +7 I Mn++ +2 "5 5

KMnO, (neutral) MnO.- +7 MnO, +4 3 3
or
Mn++++)
KMnO, (strongly MnO.- + 7. MnO,-- +6 1 1
alkaline)
~Cr.07 Cr.O,-- +6 Cr+++ +3 3' 3
HNO. (dil.) NO.- +5 NO +2 3 3
HNO. (cone.) NQ.- +5 NO. +4 1 1
CI, CI 0 Cl- , -1 1 1
Br. Br 0 Br- -1 1 1
1. I 0 r -1 1 1
Cl- 1
aHCI: IHNO.
H.O.
Cl
0
0
0 0--
-I
-2
I
2 2 -
Na.O. 0 0 0-- -2 2 2
KCIO. Cl0 3
- +5 Cl- -I 6 6
KBrO. BrO.- +5 Br- -1 6 6
KIO. 10.-- +5 r -1 6 6
NaOCI OC1-
Fe+++
+1 CI- -1 2 2 .
FeCI. +3 Fe ++ +2 1 1
Ce(SO,), Ce++++ '+ 4 Ce +++ +3 1 1
COMMON REDUCING AGENTS
H.SO. or
Na.SO. SO.-- +4 SO,-- +6 2 2
H,S S-- -2 So 0 2 2
HI 1- -1 1° 0 1 1
SnCl. Sn++ +'2 Sn++++ +4 2 2
Metals, e.g" Zn Zn 0 Zn++ +2 2 2
Hydrogen H 0 H+ +1 1 1
FeSO, (or any Fe++ +2 Fe+++ +3 1 1
ferrous salt)
Na.AsO. AsO s--- +3 AsO.--- +5 2 2
H.C.O, C.O,-- +3 C.O +4' 1 1
Ti.(SO,). Ti+++ +3 Ti++++ +4 1 1
end point the number of equivalents of the substance titrated

is equal to the number of equivalents of the standard solution
employed. We may write:
Ot NU1}1ber of gram equivalents
Normal1 y = ---:-:~.....,::.~~~--__
Number of litres
_ Number of milligram equivalents
- Number of miIliIitres (mI.)*
• See Section n, 17.
Hence: number of mg. equivalents = number of ml. X
normality. If the volumes of sol"\ltions of two different sub-
stances A and B which exactly react with one another are
VA ml. and VB m!. respectively, then these volumes severally
contain the same number 'of gram equivalents or mg. equiva-
lents of A and B. Thus:
~ X normalityA = ~ X normalitYB (i).
In practice v,., Va and "bormality A (the standard solution),
are known, hence normality B (the unknown solution) can· be
readily calculated.
Example 12. How many ml. of 0'2N hydrochloric. acid are re-
quired to neutralise 25·0 m!. of O·IN sodium hydroxide?
Substituting in equation (i), we obtain:
x X 0·2 = 25·0 X 0·1, whence x = 12·/) ml.
Example 13. How many m!. of N hydrochloric acid are required
to precipitate completely 1 gram of silver nitrate?
Th~ equivalent weight of AgNO s in a precipitation reaction is 1
mol or 169·89 g .
. Hence 1 g. of AgNO s = 1 X 1000/169'89 ='5·886 mg. equivalents.
Now number of mg.. equivalents of HCI = number of mg. equiva-
lents of AgNO a,
x X 1.= 5·886, whence x = 5·90 mI.
E:fample 14. 25 M!. of a ferrous sulphide solution,react com-
pletely with 30 m!. of 0·125N potassium permanganate. Calculate
the strength of tht! ferrous sulphate solu~ion in grams FeSO, per litre.
A normal solution of FeS04 as a reductant contains 1 mol per
litre or 151·90 g. per litre (Table IX). Let the normality of the
ferrous sulphate solution be-nA' Then:
25 x' nA = 30 X 0·125,
or nA = 30 X 0'125/25 = 0·15N.
Hence the solution will contain 0'}5 X 151·90 = 22'785 g. FeS0 4
per litre.
Example 15. What volume of 0·127N reagent is required for the
preparation of 1000 m!. of O'IN solution?
V A X normalityA = V ~ X normality B ;
VA xiH27 = 1000 X 0·1;
or VA = 1000 X 0'1/0·127 = 787·4 mi.
Hence it is necessary to dilute 787·4 mL of 0'I27N solution to I
litre. Strictly speaking it is not correct to add 212·6 m!. of water,
because there is usually a volume change on mixing. This change ,
is so small, however, that diluted solutions are often prepared by the
addition of the calculated amount of water to a measured volume
of standard' reagent. /
I, 25. Preparation of standard solutions:-If a reagent is
available in the pure state, a solution of definite normality is
prepared simply by weighing out an equivalent weight, or a
definite fraction or multiple thereof, dissolving it in the sol-
vent, usually water, and making- up the solution to a known
volume. It is not really essential to weigh out the equivalent
weight (or a multiple or sub-multiple thereof). In practice
it is often more convenient to pr~parc the solution a little
more concentrated than is ultimately required, and tllen to
dilute it with distilled water until the desired normality is
obtained. If N 1 is the required normality, VI the volume
after ,dilution, N2 the normality originally obtained and ~T2
the original volume taken, NtV l = N 2 V 2 , or VI = N 2 V 2 /N t •
The volume of water to be added to the volume V 2 is (Vl- V 2 )
ml. (compare Example 15 in Section I, 24). The following is
a list of some of the substances which can be obtained in a
state of high purity and are therefore suitable for the prepara-
tion of standard solutions: sodium carbonate, potassium
hydrogen phthalate, benzoic acid, borax, furoic acid, adipic
acid, sulphamic acid, potassium bi-iodate, sodium oxalate, silver
nItrate, sodium chloride, potassium chloride, iodine, potassium
bromate, potassium iodate, potassium dichromate and arseni-
ous oxide.
When the reagent is not available in the .pure form as in
the cases of most alkali hydroxides, some inorganic acids and
various deliquescent substances, solutions of the approximate
normality required are first prepared. These are then stan-
dardised by titration against a solution of it pure substance of
known normality. It is generally best to standardise a
solution by a reaction of the same type as that for which the
solution is to be employed, and as nearly as possible under
identical experimental conditions. The titration error and
other errors are thus considerably reduced or are made to
cancel out. This indirect method is employed for the pre-
paration of solutions of inter alia most acids (for hydrochloric
acid, the constant boiling point mixture of definite composi-
tion can be weighed out directly, if desired), sodium, potas-
sium and barium hydroxides, potassium permanganate,
ammonium and potassium thiocyanates, and sodium thio-
sulphate.
'1,26. Primary standard substances ...-A primary standard
substance should satisfy the following requirements:.
1. It must be easy to obtain, to purify, to dry (preferably at
lloo-120oq, and to preserve in a pur~ state. (This require-

ment is not usually met by hydrated substances since it is
difficult M remove surface moisture completely without
effecting partial d':composition-.)
2. It should not be hygroscopic, but should be unalterable
in ordinary ajr during weighing.
3. The substance should be capable of ,being tested for
impurities by qualitative tests of known sensitivity. (The
total amount of impurities should not, in general, exceed
0'01-0-02 per cent.)
4. It should have a high equivalent weight so that the
weighing errors may be negligible. (The pr~cision in weigh-
ing is ordinarily 0·1-0·2 mg.; for an accuracy of 1 part in
1,000, it .is necessary to employ samples weighing at least
ca. 0·2 g.)
5. The substance should be readily soluble under the con-
ditions in which it is employed.
6_ T.be reaction with the standard solution should be
stoichiometric and practically instantaneous. - The titration
error should be negligible, or easy to determine accurately by
experiment.
In practice, an ideal primary standard is difficult to obtain
and a compromise between the above .ideal requirements is
usually necessary. The substances commonly employed as
primary standards are: Acidimetry and alkalimetry-sodium
carbonate NazCO a, borax Na.2B401' potassium hydrogen
phthalate KH.C sH40 4,constantboilingpointhydrochloricacid,
potassium bi-iodate KH(IO a)2' sulphamic acid NH 2SOzOH,
succinic acid H 2 .C4 H4 0 4 • benzoic add H.C 7 H.0 2 • furoic acid
H.C5Ha02' arid adipic acid H 2 .C eH,P4'
Precipitation reactions-silver, silver nitrate, sodium
chloride, and potassium chloride.
Oxidation-reduction reactions-potassium dichromate
KllCr2 0,. potassium bromate KBrO a• potassium iodate KIO a•
potassium bi-iodate KH(IO a)2' iodine 12, sodium oxalate
Na2C2~4' arsenious oxide As 20 a,..and electrolytic, or otherwise
pure, lron.
Hydrated salts, as a rule, do not make good standards
because of the difficulty of efficient drying. However, those
salts which do not effloresce, such as borax Na 2B4 0 1,lOH 20,
oxalic acid H 2C20 4 ,2H 20 and copper sulphate CuS04.5H20,
are found by experiment to be satisfactory secondary
standards.
5§ Quantitative Inorganic Analysis
THEORY OF ACIDIMETRY AND ALKALIMETRY
I, 27. Neutralisation indicators.-The object of titrating,
say, an alkaline solution with a standard solution of an acid
is the determination of the amount of acid which is exactly
equivalent chemically to the amount of b~se present. The
point at which this is reached is the equivalence )loint, stoi-
chiometrical point, or theoretical end POint, and corresponds
to an aqueous solution of the acid and the base. If both the
acid and base are strong electrolytes, the resultant _solution
will be neutral and have a pH of 7 (Section I, 17). But if
either the acid or the base is a weak electrolyte, the salt will
be hydrolysed to a certain degree and the solution at the
equivalence point will be. either slightly alkaline or slightly
acid. The exact pH.of the solution at the equivalence point
can readily be calculated from the ionisation constant of the
weak acid or the weak .base and the concentration of the
solution (see Section I, 19). For any actual titration the
correct end point will be characterised by a definite value of
the. hydrogen ion conceptratibn of the solution, the value
depending upon the nature of the acid and the base and the
concentration of the solution. ~
A large number of substances are available, called neutrali-
sation or acid-base indicators, which possess different colours
according to the hydrogen ion concentration of the solution.
The chief characteristic of these indicators is that the change
from a predominantly" acid" colour to a predominantly
" alkaline" colour is not sudden and abrupt, but takes place
within a small interval of pH (usually about two pH units)
termed the colour change interval of the indicator. The
p'osition of the colour change interval in the PtI scale varies
widely with different indicators. For most acid-base titra-
tions, we can therefore select an indicator which exhibits a
distinct colour at a pH close to that obtaining at the equiva-
lence point.
The first useful theory of indicator action was suggested
by W. Ostwald. All indicators in general use are very weak
organic acids or bases. Ostwald (1891) considered that the
undissociated indicator acid (HIn) or base (InOH) had a
differen~ colour from that of its ion. The equilibria in aque-
ous solution may be written:
HIll ~ H+ + In- (ia) ,
and InOH ~ OH- + In+ (ib).
unionised ionil;ed
colour colour
The Theoretical Basis of Qu_antitati.ve Analysis 59
If the indicator is an anhydr~-base (In), e.g., a free amine or
substituted amine, the equilibrium is :
In + H 20 ~ OH- + HIn+ (ic).
Let us consider an acid indicator. In acid solution, i.e.,
in the presence of excess of H+ ions,' the ionis~tion will be
depressed .(common ion effe~t) and the concentra~ion.of In-
will be ·very small ;. the colour will therefore be that of the
unionised form. If the medium is alkaline, the -decrease of
[H+] will result in the further ionisation of the indicator.
[In-] increases and the colour of the ~onised form becomes
apparent. By applying the law of mass action, we obtain:
aH+ X aIn- = [H+] X [In-] X· /H+ . /In- = Kina (ii) ,
aHln [HIn] /HIn
+ [HIn]. iHIn
and [IiI ] = [In -] X Kina X IiH+' /JIn-
,.;
K
= [Unionised form] X InaX--
iHln ("')
111,
[Ionised form] iH+ '/In-
where Kina is the ionisation constant of the indicator. If

the activity coefficients are assumed to be unity-a not
entirely justifiable assumption as will be evident from the
ensuing discussion-equation (iii) reduces to the, simplified
" conct::ntration form" :
H+ - [HIn] K ~ [Unionised form] K (... )
[ ] - [I n-] X ina - . d f orm] X ina
[I OnIse Iua.
The actual colour of the indicator, which depends upon the

ratio of the concentrations of the ionised and unionised forms,
is ,thus directlY" related to the hydwgen ion concentration.
Equation (iiia) (the simplified or " classical ,: form) may be
written logarithmically:
[In-;-]
pH = log [HIli] + PKina (iv).
For a base indicator an exactly an.alogous expression to

(iii) may be deduced, which in its simplified form is :
[OH-] = [InOH] K (v),

[In+] X inb
60 Q1tafltz'tative Inorganic Analysis
where Kinb is the corresponding base dissociation constant.
This may be written:
(vi),
K inb • [InOR]
since K. = [H+] X [OH-] (approximately).
The work' of Hantzsch and others has shown that Ostwald's
simple theory of the colour change of. indicators requires
rev:ision. The newer theory, however, leads to equations
similar to the above. We can no longer regard the colour
&lange as being due to ionisation alone, for it has been shown
that an indicator consists of two or more different tautomeric
forms possessing different constitutional formulae and differ-
ent colours; one of the ·forms is a non-electrolyte (a pseudo
acid or pseudo base). the other form being an acid or a base.
The ions of the true acid or true base possess the same colour as
un dissociated form of the same structure. We can illustrate
this by reference to phenolphthalein, the chang~s for which are
characteristic of all phthalein indicators (Fig. 1-4).
HO~ ~ H.o~ () HO~ ('n

;""'-<-0" ;'~:H"- ;'~;'H'
Iy IIy lIly
OH 0 o
COlour'fess (lactone) Red (quinone) Red
(pseudo acid) (acid) (salt or ion)
Hln' ~ Hln In- +H+
FIG. 1-4.
In the solid state or in acid s()lution, it exists chiefly as the
colourless lac;tone I. This is in equilibrium in solution with a
trace of the tautomeric quinonoid form II, * which is an acid
since it possesses the carboxyl group -COOR and dissociates
to yield the R+ ion and the indicator ion III.
• The change from the benzenoid to the <tainonoid form is believed to be
responsible for the production of colour,:
. I I I I
~C-C)C-..l -;;.. C=C)
• -C,
'C
I
C
I
=C
<
C=C
I I
or more probably a state of resonance or mesomerism obtains wherein the
C=
coloured ion possesses a structure intermediate between the various possible

btmzenoid and quinonoid forms.
Consider now an acid or, more c;:orrectly, a pseudo-acid
indicator in the light of the new theory. Let HIn' represent
the pseudo acid and HIn the acid. Two equilibria must be
considered in aqueous solution:
HIn' ~ HIn (vii) ,
and HIn ~ H+ + In- ,., (viii) .
Applying the law o! mass action to both, we have:
for (vii), KT = aHln / 'aHln' (ijt) ;
and for (viii), (x).
.
The ion.. In-has the same colour as the weak acid HIn. In
acid solution, i.e., in the presence of excess of H+ ions, the
indicator will exist largely as HIn and HIn'. Now these two
forms have different colours, and HIn has the same colour as
In-. The indicator will only be useful if the equilibrium
constant for the tautomeric equilibrium (vii) is very small ;
then practically the whole of the undissociated indicator will
be in the form HIn'. Combining (ix) and (x), we obtain:
_ K K' _ -aHln aH+ X aln- ._ a....:H:_+_x_a=In_-
K ina- T X 1 - X - (xi)
aHln' aHln aHIn'
where Kina is the apparent dissociation. ~nstant of the indicator.

We have assumed that the activity (and therefore the
concentration) of HIn' to be very much greater than that
of HIn, hence aHIn' is virtually- the activity of all the
indicator not in the salt or ionic form, and the expression
(xi) is identical with (ii) deduced from the Ostwald theory.
Equation (xi) may be written in the following forms: •
K "_ [H+] X (In-] fH+ X fIn-
(xia) ;
ina - [HIn'] - ,X fHIn'
.. ~ + [H+] = [HIn'] X K- X fHIn'

:t H [In-] ma];n _
[Form .with " acid" colour] K' fHIn'
=---~~ ___-:---:-:----::,.---= X I X - (xii).;
[Form with "alkaline" colour] . na fln-
or pH = log [In-] / [HIn'] + pKina + log fIn- / !HIn"

= log [In~] / [HIn'] + pKina + loghn- (xiii),
since pH = - log fH+ [H+] and fHIn' may be regarded as
constant. .
62 Quantitative Inorganic. Analysis
If the .activity coefficient of the indicator anion may be
expressed by the usual equation (Section 1,20)", it. follows that:
pH = pKina + log [In-] / [HIn'] - 0·505 Zj2 I-' 0·5 + Cp..(xiiia).
In dilute solution (I-' ::f> 0'01) the term CI-' may be neglected,
and the equation reduces to :
pH = pKina + log [In-] / [HIn'] - 0·505 Z';2fLo.s (xiiib).
The colour of an indicator is determined by the ratio of the
concentrations of HIn' andJn - (the pseudo acid and the ionised
form), but the actual pH will depend upon the ionic strength
of the solution (equation xiiib). JIence when making a colour
comparison for the determination of the pH of a solution not
only must the indicator concentration be the same in the two
solutions, but the ionic strengths must be equal or approxi-
mately, equal. Equations (xiiia) and (xiiib) incidentally
provide an explanation of the SOc called " salt effect" which is
observed with indicators.
For many practical purposes it is sufficient to write equation
(xiiia) in the form:
pH = pK~a + log [In-] f [HIn'] (xiiic),
provided the influence of ionic strength be.. borne in mind.
As already mentioned the colour of an indicator is determined
by the ratio of the concentrations of the" acid" and" alka-
line" forms.. Both forms are ptesent at any hydrogen ion
concentration. It must be realised, however, that the human
eye has a limited ability to detect either of the two colours
when one of them predominates. Experience shows that the
solution will.appear to have the" acid" colour, i.e., of HIn', up
to the ratio of [HIn'] to [In-] of .approximately 10, and the
" alkaline" colour, i.e., 01 In-, up to a similar ratio of [In-]
to [HIn'] of approximately 10. Thus only the "acid"
colour will"be visible when' [HIn'] / [1n-] ,< 10, the corres-
ponding limit of PH given by equation (xiiic) is:
pH = pKina - 1 ;
only the "alkaline" colour. will be visible when
.[In-] / [HIn'] < 10, and the corresponding limit of PH is :
pH = PKlna + 1.
The colour change interval is 'accordingly pH = pKina ±I,
i.e., over approximately two pH units. Within this range the
indicator will appear to change f:(om one colour to the other.
The change will be gradual, since it depends upon the ratio
of the cpncentrations of the two coloured forms (pseudo acid
and ionised). When thE; pH of the solution is equal to the
apparent dissociation constant of the indicator PKina , the
ratio [HIn'] to [In-] becomes equal to 1, and the indicator will
have a tolour due to an equal mixture of the" acid" and
" alkaline" forms. This is sometimes known as the middle
tint of the indicator. This applies strictly only if the two
colours are of equal intensity. In one form is more intensely
coloured than the other or if the .eye is more sensitive to one
colour than the other, then the middle tint will be slightly
displaced along the pH range of the indicator. ..
For an indicator base we have, according to the modified
theory, the two equilibria:
In'OH "'" InOH (xiv),
InOH "'" In+ + OH-
and
. (xv).
.
The following equation may be deduced along similar lines
to that described for an indicator acid:
- [In'OH] jIn'oR
/OH- [OH.1 = [In+] X K infJ X /In+
[Fotm with" alkaline" colour] K fIn'oH
= ' X 'b X -- (xv)
[Form with " aci~ " colour] ill frn+ '
where Knb is the apparent dissociation constant of the base.
Remembering that Kw = fH+ [H+] X jOH- [OH-], we may ~
write: .
K [In+] I +
!R+ [H+] = __ w X X ~ (xvi).
K inb [In'OH] jIn'OR
The equation is of the same form as (xii) since the ratio of the
two constants may be replaced by a new constant K inb,.
The above theory applies ohly'to a monobasic acid or to a
monoacid base. Complications arise if two such changes are
poss~ble, and the transition range may be increased.
The apparent ionisation constant of an indicator may be
determined approximately by finding .either the PH of a
solution (Section I, 17) containing it when it shows its middle
tint, or by finding the range of hydrogen ion concen'tration in
which definite colour changes are observed; the pH of such a
solution lies within the limits PKina (or pKinb ) ± 1. A
.more accurate method depends upon the determination of the
fraction of the indicator (e.g.,' [In-] / [HIn']) which has
changed colour in a series of solutions of known pH (buffer
solutions-:-Section I, 20); the approximate apparent dis-
64 Quantitative Inorganic' Analysis
sociation -constants may then be calculated from equations
(xii) or (xvi), the activity· coefficients being assumed to be
Jmity. With a one colour indicator the colour intensity is
measured, e.g., by means of a colorimetric or spectrophoto-
metric method, in a series of solutions of equal concentration
ot indicator and different values of pH. .J'or an indicator of
the phenolphthalein type, the standard of Fomparison will
be a solution containing an appreciable 'excess of alkali:
this will corresp~nd to a value 9f [In-] equal to the total
indicator coru:entration. With a two-colour indicator, various
modifications must be introduced, which cannot, however,
be discussed here. In some cases, e.g., p-nitrophenol and
methyl orange, the constants may be determined by conduct-
ance measurements or by a study of the hydrolysis of salts.
Some approximate values of the apparent dissociation con-
stants of indicators, which we may term indicator constants
Kin are given in Table X. This Table contains a sele~ted
list of indicators suitable for volumetric analysis and also for
the colorimetric determination of pH (Section V, 6).
TABLE X. COLOUR CHANGES AND pH RANGE OF CERTAIN INDICATORS.
-
Colour Colour
Indicator Chemical Name pH in Acid in Alka- pKln
Range Solution lin~
- ,
Solution
•
Brilliant cresyl Amino-diethyl- 0·0- 1·0 Red- Blue
blue (acid) amino-methyl d.i- orange
phenazonium
chloride
Cresol red (acid) a-Cresolsulphone- 0·2- 1·8 Red Yellow
phthalein
Quinald.ine r~ a-(p-dimethyl-
amino - phenyl- ),0- 2·0 ColourlesS Red
ethylene) - quino-
line ethiodide
-
.
Thymol blue Thymol-sulphone- ).2- 2·8 Red Yellow 1·7
(acid) phthalein
Tropaeolin 00 Diphenylamino - p 1·3- 3·0 Red Yellow
benzene - !\odium
sulphonate
Meta-cresol m - Cresol - sul- 1·2- 2·8 Red Yellow
purple phone-phthalein
Bromo-phenol Tetrabromophenol- 2·8- 4·6 Yellow Blue 4·0
blue sulphone - phtha-
lein
Methyl yellow Dimethylamino- 2·9,- 4'0 Red Yellow
I
. Methyl orange
azo-benzene
Dimethylamino-
azo-benzene so-
3·)- 4·4 Red . Yellow 3·7
d.ium sulphonate •
TABLE X-continued.
-
Colour Colour
Indicator Chemical Name pH in Acid in Alka- PKi{"
Ran9f Solution line
Solution
Congo)'ed Diphenyl - dis - 3·0- 5·0 - Violet Red
azo - a - naphthy-
lamine - 4, - suI-
phonic acid
Bromo-cresol Tetrabromo - m - 3'8- 5·4 Yellow Blue 4'7
green cresol - sulphone-
phthalein
Methyl red o-Carboxybenzene- 4·2- 6·3 Red' Yellow 5-l
azo - dimethyla-
niline
Chloro.phenol Dichloro-phenol- 4·8 -6·4 Yellow Red 6·0
red sulphone - phtha-
lein
p-Nitrophenol p-Nitrophenol 5·6- 7·6 Colourless Yellow H
Bromo -cresol Dibromo-o-cresol- 5·2- 6·8 Yellow Purple 6·3
purple sulphone - phtha-
lein
Azolitmin - 5,0- 8·0. Red Blue
(litmus)
Bromo-thymol Dibromo - thymol- 6·0- 7·6 4 Yellow Blue 7-0
blue sulphone - phtha-
lein
Neutral red Amino - dimethyl- 6·8 -8'0 Red Orange
amino - tolu -
phenazonium
chloride
Phenol red Phenol - sulphone- 6·8- 8'4 Yellow Red 7·9
phthalein
Cresol red (base) o-Cresol-sulphone- 7·2- 8'8 Yellow Red 8·3
phPlalein
a-Naphthol- a-Naphtholphtha- 7·3- 8'7 Yellow Blue 8'4'
phthalei].; lein
Thymol blue Thymol-sulphone- 8·0- 9·6 Yellow Blue 8·9
(base) phthalein
Phenolphthalein Phenolphthalein 8·3-10·0 Colourless Red 9·6
Turmeric - 8·0-10'0 Yellow Orange
Thymolphtha- Thymolphthalein 9·3-10·5 Colourless Blue 9·2
lein
Alizarine p - Nitrobenzene· iO·I-12·0 Yellow Orange
yellowR azo - salicylic red
acid
Brilliant cresyl (see above) 10·8-12'0 Blue Yellow
blue (base)
Tropaeolin 0 p - Sulphobenzene- 11·1-12'7 Yellow Orange
azo-resorcin.ol
Nitramine 2 : <1 : 6-Trinitro- 10·8-13'0 Colourless Orange-
phenyl - methyl -
nitroamine . Brown
It is necessary' to draw attention to the pH of various types

of water which may be encountered in quantitative analysis.
Water in equilibrium with the normal atmosphere containing
D
0·03 per cent by volume of carbon dioxide has a pH of about
5'7; very carefully prepared cohductivity water has a pH
close to 7; water saturated 'with carbon dioxide under a
pressure of one atmosphere has a pH of about 3·7 at 25°C.
The analyst may therefor~'oe dealing, according t9 the con-
ditiqns that prevail in the laboratory, with water having a
PH between the two extremes pH 3·7 to pH 7. Hence for
indicators which show their alkaline colours at pH values
above 4'5, the effect of carbon dioxide introdu'Ced during a
titration either from the atmosphere -or from the titrating
solutions must be seriously considered. This subject is
discus!'/ed again later (Section I, 32).
The colour change intervals of most of the various indicators
listed in the Table are represented graphically in Fig. 1-5.
Brilliant cresyl blue RO~...?I B BI ~ IY ,

aresol red Ij~~ I r~~~","1
•
Quinaldine red C ~ IR
Thymol bl•• Y~ ~B
Tropaeolm 00 I'l.:< IY
lfeta-cre,Q/_IlIIrpl.
If.thylJlel~
lIelhyl olYIng'
'blu,_
- YR~rB
R ~Y
. B
C
o
- 8111.
=
Colour/'"
= Orang.
08= (/"'Qnge-brewn
~
r-
Oongo red P =
Purple
-
~
R =Rsd -
or.mo-crosol green Y~
RO= Red·orange -
Methyl red
red
Para-nitro phen.1
y
Y =re/lo",'
-
, purple Y!2
, blu.
Neutral red
Pheno/~d_ Y ~
«-Naphthol phthalein
Ph.nol phthalei"
:12Z:
~ ~ lORI(
Thymol phthalem
Alizarine ye/low R
Trapaeolin 0
Nltromine
0 1 2 3 4
, (; 6 7 B 9 11 11 12 13
pH units
FIG. 1-5
It will be seen that the pH range between 0 and 13 is covered

by the above indicators.
The indicators at present most used iIi quantitative analysis
are methyl orange (as the simple or "screened" forms-.
see Section I, 29), methyl red and phenolphthalein. The
phthalein indicators merit much more attention than has
hitherto been given to them; this wou,ld be the case if their
many advantages, particularly their sharp col'Our changes
The Theoretical Basis of Quant£tative Analysis 67
and slight sensitivity to carbori dioxide, were more .widely
realised. .
I, 28. Preparation of indicator solutions.-As a rule the
stock solution of the indicators contain 0,5 to I g. of. indicator
per litre of solvent. If the substance is soluble in water,
e.g., if a sodium salt, water is the solvent; in most other cases
70 to 90 per cent alcohol is employed. It should now be
stated that the new synthetic indicators, particularly the
sulphonephthaleins and phthaleins which exhibit brilliant
colour chang~s, may be used with confidence in all those cases
where the older ones, largely natural products, were formerly
employed.
Methyl orang~. This indicator is encountered in commerce
either as the free acid or as the sodium salt.
Dissolve 0·5 g. of the free acid in 1 litre of water, and filter
the cold solution if a precipitate separates.
Dis!iolve 0·5 g. of the sodium salt in 1 litre of water, add
15·2 in!. of O'IN hydrochloric acid, and fiiter, if necessary,
when cold.
Methyl red. Dissolve 1 g. of the free acid in 1 litre of hot
water, or dissolve in 600 ml. of alcohol and dilute with 400
ml. of water.
Phenolphthalein. Dissolve 5 g. of the reagent in 500 ml.
of alcohol and add 500 m!. of water with constan~ stirring.
Filter, if a precipitate forms.
Thymolph'thalein. Dissolve 0'4 g. of the r~agent in 600 ml.
of alcohol and add 400 ml. of water with stirring.
a-Naphtholphthalein. Dissolve 1 g. of the indicator in 500
ml. of alcohol and dilute with 500 ml. of water.
Litmus. Commercial litmus contains, in addition to the
colouring matter (azolitmin), varying quantities of calcium
carbonate, calcium sulphate, etc. Purification is effected as
follows. Digest 10 g. of the commercial substance with 35
rol. of rectified spirit on the water bath for 1 hour and decant
the alcohol; repeat this process 'twice so that 105 m!. of
alcohol are employed in all. Ext.ract the residue several times
with water (total volume 175 ml..,-the fi];st 25 ml. is discarded
as it contains most of the potassium carbonate), and allow to
stand for several days. Decant or siphon off the clear ex-
tract. This is of suitable concentration for most practical
purposes.
As a general rule the use pf litmus in quan.titative neutral-
isations should be avoided, for the solution may contain
68 Quantitative, Inorganic Analysis .
variable amounts of the colouring matter, It is useful for
qualitative tests, but then litmus paper is to be preferred to
the solution. Such indicators as bromo-cre~l purple or
bromo-thJ:mol blue, which cover' approximately the same pH
range, are excellent substitutes. for litmus; the colour changes
are very striking and a pure product can be obtained, 'so that
a uniform solution is assured.
Azolitmin. This is the pure litmus colouring matter ·and
consi~ts largely of the blue potassium salt of (red) azolitmic
acid. A solution is prepared by dissolving 1 g. in 1 litre of
water. It is not very much more .sensitive than good litmus
solution.
Sulphonephthaleins. These indicators ·are usually supplied
in {he acid form. They are rendered water soluble by adding
sufficient sodium hydroxide to neutralise the potential sul-
phonic.acid group. 1 gram of the indicator is triturated in a
-clean glass mortar with the appropriate quantity or O·lN
sodium hyaroxide solution, and then diluted with water to 1
litre. The following volumes of O·lN sodium hydroxide are
required for 1 g. of the indicators: bromo-phenol blue, 15·0
ml.; bromo-cresol green, 14·4 ml.; bro~o-cresol purple,
18·6 ml.; chloro-p'henol red, 23·6 ml.; bromo-thymol blue.
16·0 ml.; phenol red, 28·4 mI.;. thymol blue, 21·5 ml. ;
cresol red, 26·2 ml.; meta-cresol purple, 26·2 ml.
Quinaldinic acid. Dissolve 1 g. in 100 ml. of rectified
spirit.
Methyl yellow,' neutral red, and congo red. Dissolve 1 g.
of the indicator in 1 litre of 80 per cent alcohol. Congo red
may also be dissolved in water.
P-Nitrophenol. Dissolve 2 g. of the solid in 1 litre of'water.
Alizarine yellow R. Dissolve 0·5 g. of the indicator in 1
litre of rectified spirit.
Tropaeolin 0 and Tropaeolin 00.' Dissolve 1 g. of the solid
in 1 litre of water.
Many of the indicator solutions are available commercially
already prepared for use. These should be bought from the
actual chemical manufacturers, who will usually supply full
details as to the method 'of preparation, concentration of the
solution, etc.
I. 29. Mixed indicators.-With some indicators, e.g.,
methyl orange, the colour change is not easily detectable,
partiq.llarly in artificial light .• A sharper and more pro-
nounced colour change may often be obtained by using a
suitable mixture of the indicator and an indifferent dye.
Thus for methyl orange, Moerk (1921) used a mixture of 1 g.
of methyl orange and 2·5 g. of indigo carmine dissolved in 1
litre of water; the colour change from the alkaline to the
acid side is yellowish-green -+ grey -+ violet, the grey stage
being reached at pH 4. Hickman and Linstead (1922)
obtained a somewhat better effect by dissolving a mixture of
1 g. of methyl orange and 1·4 g. -of xylene cyanol FF* in
500 ml. of 50 per cent alcohol; here the colour change from
the alkaline to the acid side is green -+ grey -+ magenta, the
middle grey stage being obtained at pH 3·8. The ~bove is an
example of a.screened indicator, and The solution is sometimes
known as " screened" methyl orange.
For some purposes it is desirable to have a sharp colour
change over a narrow and selected range of hydrogen ion
concentration; this is not easily seen with an ordinary indi-
cator since the colour change uS"!lally extends over two -units
of pH. The result may, however, be achieved by tne use of a
suitable mixture of indicators. Four examples will be given.
(a) A mixture of equal parts of neutral red (0'1 % solution
in alcohol) arid methylene blue (0·1 % solution in alcohol)
gives a sharp colour change from violet-blue to green in pass-
ing- from acid to alkaline solu~ion at pH-7. This indicator.
may be employed to titrate acetic acid with ammonia solution
or vice versa. Both acid and base are approximately of the
same strength, hence the equivalence point will be at a pH
of ca. 7 (Section I, 35); owing to the extended hydrolysis and
the flat nature of the titration curve (see Section 1, 35), the
titration cannot be performed except with an indicator of
very narrow range.
(b) A mixture of equal parts of methyl orange or of lllethyl
yellow (0·1 % solution in alcohol) and methylene blue (0·1 %
solution in aloohol) may be used in the titration of the very
weak base pyndine and a strong acid; in this case also the
* The constitution is :
CHi
/
/OH /O-NHC1H.
NaO,S -O-c "-

"'-OH I "-()-NHC.H.
OH ~
CHa
titration curve is very flat. This indicator is also useful for
other somewhat stronger bases and strong', acids.
(C) A mixture of phenolphthalein (2 parts of 0·1 % solution
in 50% alcohol) and a-naphtholphthalein (1 part of 0·1%
solution in 50% alcohol) passes from pal~ r,9se through green
to violet at PH = 9·6. The mixed indicator is suitable for
the titration of phosphoric acid to 'the dibasic stage (K2 =
6·2 X 10-8 ; equivalence point at PH = ca. 9·7).
Phenolphthalein, which was previously useq., is not so satis-
factory since its pH range is 8·3 to 10 and the titration curve
is relatively flat (Section I, 36).
(d) A mixtur.e of thymol blue (6 parts of 0.:.1% aqueous
solution) and cresol red (1 part of 0'1% aqueous solution)
and cresol red (1 part of 0·1 % aqueous solution) is violet at
pH 8·4, blue at pH 8'3, and rose at PH 8·2. It has been
recommended by Simpson (1924) for the titration of carbo~ate
to the bicarbonate stage.
I,OO. Universal"or multiple range indicators.-By suitably
mixing certain indicators the colour change may be made to
extend over a cOBsiderable porti~ of the P!i range. Such
mixtures are usually called "universal. indicators." They
are not suitable for quan~itative titrations, but may be
employed for the qetermination of the approximate PH of a
solution by the colorimetric method. Bogen (1927) gives the
following details: dissolve 0·1 g. of phenoltJhthalein, 0·2 g.
of methyl red, o·a g. of methyl yellow, 0·4 g. of bromo-
thymol blue, and 0"5 g. of thy'r1lo1 blue in 500 m!. of absolute
alcohof, and add sufficient sodium hydroxide solution until the
colour is yellow. The colour. changes are as follows: pH 2,
red; PH 4, orange; pH 6, yellow; pH 8, green; pH 10, blue.
The following recipe is due to T. B. Smith (1929). Dis-
solve 0·05 g. of methyl orange, 0·15 go, of methyl red, 0·3 g. of
bromo-thymol blue, and 0·35 g. of phenolphthalein in 1 litre
of 66 per cent alcohol. The colour changes are: pH up to 3,
red.; pH 4, orange-red: pH 5, orange; pH '6, yellow; pH 7,
yellowish green; PH 8, greenish-blue; PH 9, blue; pH 10,
violet; pH 11, reddish:violet. Several" universal indica-
tors" are available commercially: that-of the British Drug
Houses, London, is most widely used.
Colorimetric determination of pH. This is discussed fully in
Chapter V. .
I, 31. Neutralisation curves.-An iJtsight into the mechan-
ism of neutralisation processes is obtained by stu<l¥ing the
changes in the hydrogen ion concentration during·the course
of th.e appropriate titration. The change in P:tI in the neigh-
bourhood of the ~quiva1ence point is of the greatest import-
a~ce as it enables us to select an indicator which will give the
smallest titratioh error. The curve· obtained by plotting pH
as ordinates against the percentage of acid neutralised (or the
number of ml. of alkali added) as abscissae is known as the
neutralisation (or, more generally, the titration) curve. This
may "e. determined experimentally by an estimation of the
PH at various stages during the titration by an electrometric
method (Section 1,52), or it'may be computed approximately
with the aid of the theoretical principles that we have already
studied. We shall, for the present, adopt the latter method.
1,32. Neutralisation of a strong acid and a strong base.-
We shall assume that both the acid and the base-are completely
dissociated and that the activity coefficients of the ions. are
unity in order to calculate the change ofpH during' the course
of the neutralisation of the strong acid and the strong base,
or vice versa, at the laboralory temperature. For simplicity
of calculation we 'shall start with 100 ml. of, say, N hydro-
chloric acid and add N sodium hydroxide solution. The pH
of N hydrochloric acid is O. When 50 ml. of the N base
have been added, 50 ml. of un-neutralised N acid will be
present in a total volume of 150 ml. ; ,
[H+] will therefore be 50 X 1/150 = 3'33 X 1.0-1, or
pH = 0·48.
For 75 ml. of base, [H+] = 25 X 1/175 - 1·43 X 10-1 ,
PH = 0·94.
For 90 ml. of base; [H+] = 10 X 1/190 5·27 X 10- 2 ,
pH = 1·3.
For 98 ml. of base, [H+] = 2 X 1/198 - 1·01 X 10- 2 ,
pH = 2·0.
For 99 ml. of base, [H+] = 1 X 1/199 - 5'03 X 10-3 ,
pH = 2·3.
For 99·9 mL of base, [H+] = 0·1 X 1/199·9 . 5·01 X 10- 4 ,
pH = 3·3.
Upon the addition of 100 mI. of base, the pH will change
sharply to 7, the equivalence' point, provided carbon dioxide
is absent; the resultant solution is equivalent to one of
sodium chloride.
With 100'1 ml. of base, [OH-] = 0'1/200'1 = 5·00 X 10-4,
pOH = 3·3 and pH = 10'7,
With 101 ·ml. of base, [OH-] = 1/201 = 5·00 X 10-3 ,
POH = 2·3, and pH = 1l.7.
These results show that as the titration proceeds, the PH
rises slowly, but between the addition of 99·9 m!. and 100'1
inl. of alkali, the pH of the solution rises from 3·3 to 10·7,
i.e., in the vicinity of the equivalence point the rate of change
of pH of th.e solution is very rapid.
The complete results, extended to 200 m!. of alkalj, are
collected in Table XI; this also includes the figures for
0'1N and o'OlN solutions of acid and base respectively. The
additions of alkali have been extended in all three cases to
200 m!.; it is evident that the range from 200 ml. to 100.ml.
and bey'onel represents the reverse titration of 100 ml. of
alkali with. the acid in the presence of the non-hydrolysed
sodium chloride solution. The data in the Table are presented
graphically in Fig. 1-6:
TABLE XI. pH DURING TITRATION OF 100 ML. OP Hel WITH NaOH OP

. EQUAL NORMALITY.
MI. of NaOH. N solution O'IN solution ~ O'OIN solution

added pH pH pH
I I
- 0 - 0·0 1·0 2·0
50 0·5 1'0 2·5
75 0:8 1·8 2·8 •
90 1'3 2·3 3·3
98 2'0 3·0 4,·0
99 2'3 3·3 4·3
99'0 2·6 3·6 4·6
99·8 3·0 4·0 5'0'
99·9 3·3 4'3 5·3
100·0 7·0 7·0 7·0
100·1 10·7 9·7 8'7
100·2 11·0 10·0 9·0
100·5 11·4 10·4 9·4
101 11·7 10·7 9·7
1()2 12·0 H'O 10·0
llO 12'7 11-7 10·7
125
'150
. 13·2
13·5
12·2
12'0
11·2
11·5
200 14·0 13·0 12·0
In quantitative analysis we are especially interested in the

changes of. pH near the equivalence point. This part of Fig.
1-6 is accordingly shown on a larger scale in Fig. 1-7, on which
are also indicated the colour change intervals of some of the'
common indicators.
14 14
~
13 ~~ 13
I ~~ ,_
~
12 12
~~
~
If ().
rf
J I
10 Io
8 8
pH pH
7 7
a
Il 6
4 4
3 /
~
O· ~ v J. 3
2 2
~ V
~
"""'" ~ ........ I
I-- ~
00
""""20 40 60 80 100 120 140 150 lBO 208
Alkali added. mi.
FIG. 1-6. Neutralisation Curves of 100 ml. of HCI with NaOH of same
Normality (Calculated).
With N solutions, it is evident that any indicator with an
effective range between pH 3 and 10·5 may be used. The
colour change will be sharp and the titration error negligible.
With. O'IN solutions, the ideal pH range for an indicator is
limited to 4'5-9'5. Methyl orange will exist chiefly in the
alkaline form when 99·8 mI.· of alkali have been added and the
titration error will be 0·2 per cent which is negligibly small
for most practical purposes; it is therefore advisable to add
sodium hydroxide solution until the indicator -is present
completely in the alkaline form. The titration error is also
negligibly small with phenolphthalein.
With O'OIN solutions, the ideal pH range is still further
limited, to 5'5-8'5; .such indicators as methyl red, bromo-
thymol blue or phenol red will be suitable. The titration
error for methyl orange will be I to 2 per cent.
The above considerations apply to solutions which do not
contain carbon dioxide. In practice, carbon dioxide is
usually present (compare Section I, 27); it may be derived
from the small quantity of carbonate in the sodium hydroxide
D*
74 Quantitative Inorganic A nalY6is
14
~.
, I-- ~
I
/'
1./ ~ ~
~ p-'""
,,.
8-- Phenol pbtba1da
,
1"""'
In. blue
~ Methyl red
4 ~
-
~~ f..- J
V
V
J
J
~~
~ Methyl oran,e
.}Ji. ~ V
1
.. Indicator
Ranges
0
98 99 100 101 102
Alhali added. mi.
FIG. 1-7. Neutralisation Curves of 100 mI. of HCl with NaOH of same
Normality in Vicinity of Equivalence Point (Calculated).
and/or from the atmosphere. The gas is in equilibrium with
carbonic acid, of which both stages of ionisation are ,weak.
This will introduce a small error when indicators of high pH
range (above PH 5) are used, e.g., phenolphthalein or thymol-
phthalein. More acid indicators, such as methyl orange and
methyl yellow, are unaffected by carbonic acid. Kolthoff
has calculated that the difference in .the amounts of sodium
hydroxide solution used with methyl orange and phenol-
phthalein_is not greater than 0'15 to 0·2 ml. of O·lN sodium
hydroxide when 100 ml. of O'lN hydrochloric acid are titrated.
A method of eliminating this error, other than that of selecting
an indicator with a pH range below pH 5, is to boil the solution
while still acid and to continue the titration with the cold
solution. Boiling the solution is particularly efficacious when
titrating dilute (e.g., 0'01]\1') solutions.
I, 33. Neutralisation of a weak acid with a strong base.-
We shall confine our attention to O'lN solutions; other
concentrations can be treated analogously. Let us study the
neutralisation of 100 ml. of O'lN acetic acid with O'IN sodium
The Theoretical Basis of QuantitaJive Analysis 75
hydroxide solution. The PH 6f the solution at the equival-
ence point-is given by (Section I, 19) :
pH = 1/2 PKw + 1/2PKa + 1/2 log c,
= 7 + 2·37 + 1/2 (-1) = 8·87.
For other concentrations, we may employ the approximate
mass action expression:. •
[H+1 X [CIlH aOIl-1/ [H.CIlHaO Il] -:- K" {i),
or [H+] = [H.CIlHaOJ X K., / [CIlHaOIl-J,
or pH = log [Salt] / [Acid] + pKa (ii).
The concentration. of the acid at any point is taken as equal
to the concentration of tlie acid that has not yet been neu-
tralised, and the concentration of the salt is assumed equal
to the concentration of alkali added. These approximations
are not valid at the beginning of the neutralisation, but since
we are interested only in the pH change near the stoichio-
metric point, the subsequent argument is unaffected.
The initial PH of O'IN acetic acid is computed, from equa-
tion (i); the dissociation pf the acid is relatively so small that
it may be neglected in expressing the concentration of acetic
acid. Hence from equation (i) :
[H+] X [CIl1I30Il-] / [H.CIlHaO Il] = 1·82 X 10--5,
or [:fI+)2 / 0·1 = 1'82 X 10-5,
'or [H+] = vl'82 X 10- 8 = 1·35 X 10-3 ,
or pH = 2·87,
When 50 ml. of O'lN alkali have been added,
[Salt] = 50 X 1·1/150 = 1·33 X 10-11,
and [Acid] = 50 xO·l/150 = 1·33 X 10-2 ,
pH == log (1·33 X 10-2 /1.33 X 10-2) + 4·74,
= 4·74:. .
The pH values at other points on the titration curve are
similarly calculated. After the equivalence point has been
passed, the solution contains excess of OH- ion~ which will
repress the hydrolysis of the salt; the pH may be assumed,
with sufficient accuracy for our purpose, to be that due to
the excess (}f base present so that in this region the titration
curve will almost coincide with that for O'lN hydrochloric
acid (Fig. 1-6 and Table XI). All the i'esult~ are collected in
Table XII, and are depicted graphically in Fig. 1-8. The
76 Quantitative
, Inorganic Analysis
results for the titration of 100 ml. of 0'1N solution of a weaker
acid (K" = 1 X 10-7) with O·lN sodium hydroxide at the
laboratory temperature are also included.
TAllLE XII. NEUTRALISATION 01' 100 ML. OF O'lN ACETIC ACID (KA =
1·82 x 19- 6) AND OF 100 ML. OF O·lN HA (K.. = 1 x 10-') WITH O·lN SODIUM •
HYDROXIDE
,----
O'IN Acetic Acid O'IN HA (Ka =1 X lO-')
MI. 010'1 N NaOH used pH pH
0 2'9 4'0
10
25
50
90
3'8
4'3
4·7
5·7
. 0·0
0'5
7·0
8'0
99·0 6'7 9'0
99·5 7'0 9'3
99·8 7'4 9'7
99·9 7·7 9·8
100·0 8'9 10'0
:-p
(l~ ~
12
I
II
.~~.~:~ phthalein
-~~.;: ~.of(
9 .,' ~. phthalein
~1J ,.- ~
8
pH
7~
i\,'l;; ~
6 r:;') ~
5:( IMethyl red
4
l,.J: ~ ~
~MethYI orange
~ ~
1/
Indicator
Ranges
0
0 20 -40 60 -80 1()O 1'10 149 160 SO 200
AIHal1 aWed, mI.
FIG. 1-8. jl{eutralisation Curves of O'IN Acetic Acid and of O·lN Acid
(Ka = I x 10-7) with O'IN Sodium Hydroxide (Calculated).'
For O'IN acetic acid and O'IN sodium hydro"ide, it is evident
from the titration curve that neither methyl orange nor methyl
red can be used as indicators. The equivalence point is at
PH 8'9, and it is necessary to use an indicator with a pH
range on the slightly alkaline side, such as phenolphthalein,
thymolphthalein or tliymol blue (PH range, as base, 8·0-
9·6). For the acid with K" = 10-7 , the equivalence point is
at pH = 10, but here the r~te of change of PH in the neigh-
bourhood of the sooichiometric point is very much less pro-
nounced owing to considerable hydrolysis. Phenolphthalein
will commence to change colour after about 92 ml. of alkali
have been a.dded and this change will occur to the equivalence
point; #thus the end point will not be sharp and the titration
error will be appreciable. With thymolphthalein, however,
the colour change covers the PH range 9·3--,--10·5 ;., this in.di-
cator may be used, the end point will be more sharp than for
phenolphthalein, but nevertheless somewhat gradual, and the
titration error will be about 0·2 per cent. Acids that have
dissociation constants less than 10-7 cannot be satisfactorily
titrated in O'IN solution with a simple indicator.
In general, it may be stated that weak acids (Ku > 5 X 10-')
shQuld be titrated with phenolphthalein, thymolphthalein, or
thymol blue as indicators. .
I. 34. Neutralisation of a weak base witn a strong acid.-
We may illustrate this case by the titration of 100 ml. of
O'IN ammonium hydroxide (Kb = 1·8 X 10-5 ) with O'lN
hydrochloric acid at the ordinary laboratory temperature.
The pH of the solution at the equivalence point is given by
the equation (Section I. 19) :
pH = 1/2Kw - 1/2 PKb - 1/2 log c,
7 - 2·37 - 1/2 (- 1) = 5·13.
For other concentrations, the pH may' be calculated in the
following manner. Using the approximate mass action
expression:
[NH,+] X [OH-] / [NH 40H] = Kb (i),
or [OH-] = [NH,OH] X Kb / [NH4+] (ii),
or POH = log [Salt] / [Base] + PKb (iii),
or pH = pKw - PKb - log [Salt] / [Base] (iv).
The concentration of the base at any point is assumed equal
to the concentration of the base that has not been neutralised,
and the concentration of the salt is taken as. equal to the
concentration of"acid 'added,. These approximations are not
valid at the beginning of ,the neut'ralisation, but the error
thus introduced is small. After the equivalence point has
been reached, the solution contains excess of H+ ions, hydro-'
lysis of the salt will be repressed, and the subsequent pH
changes may be assumed, with sufficient accuracy for our
purpose, to be those due to the excess of acid present so that
in this region the titration curve will almost coincide with
that for O'IN sodium hydroxide (Fig. I-6ta.nd Table XI).
The results computed in the above manner are given in
Table XIII, and are represented graphically in Fig. 1-9. The
results for the titration. of 100 m!. .of a O·IN so1ution of p.
weaker base (Kb = 1 X 10-7 ) are also included. •
TABLE XIII. NEUTRALISATION OF 100 ML. OF O·lN AMMONIUM HYDROXIDE
(Kt, = 1·8 x 10-') and of 100 ML. OF O·IN BOH (Kb = 1 X lO-') with O·IN
HYDROCHLORIC ACID.
- O'IN Ammonium O·IN BOH

MI. ofO'IN HCI added Hydroxide (Kb = 1 X 10"")
pH pH
0 11-1 10·0
10 10'2 8·0 •
• 25 9'8 7'0
.50 9'3 7·0
90 8'3 6·1'
99·0 '7'3 5·0
99'5 7·0 4·7
99·8 6·6 4'3
99·9 6'3 4·0
100'0 5-l 4·0
It is clear that neither thymolphthal~in nor phenolphthalein

can be employed in the titration of O'IN ammonium hydrox-
"ide. The equivalence p~int is at pH 5'1, and it is necessary
to use an indicator with a pH range on the slightly acid side
(3-6·5), such as methyl orange, methyl red, bromo-phenol
blue or bromo-cresol green.· The last-named indicators may
be utilised for the titration of all weak bases (Kb > 5 X 10-1)
with strong acids.
For the weak base (Kb = I X 10-7), bromo-phenol blue or
methyl orange may be used; no sharp colour change will be
obtained jVith bromo-cresol green or with methyl red, and the
titration -error will be consideraple. .
I, 35. Neutralisation of a weak acid with a weak base.-
This case is exemplifiea. by the titration of 100 ml. of O·lN
14
32
.
:&
~
~ Thymol phthalein
~
a\ ~ I.&~, ~ Phenol phthalein
,~
.u
~
pH ~.~ \
7
~
~
5
\ Methyl red
.
4
• orange
-
3
\;.
2
.--... ~ .!!£!
1 Indicator
Ranges
n
2 ~o 6 o E10 ~10 1: 10 140 10 'II 10 21 )0
Alhali added. mi.
FIG. I-g. Neutralisation Curves of 100 ml. O'lN Ammonium Hydroxide

(Kb'= 1'·8 x 10-') and of O'lN Base (Kb = 1 x 10-') with O'1N Hydrochloric
.. Acid.,
ac~tic acid (Ka --;- 1·8 X 10-5) with O·IN ammonium

hydroxide (Kb = 1'8 X 10-5). The pH at the equivalence
point is given by (Section I, 19) :
pH = 1/2 PKUI + 1/2 pKa - 1/2PKb ,
- 7·0 + 2·37 - 2·37 = 7·0.
The calculation of the neutralisation curve is a iittle more
difficult. * The titration curve for the neutralisation of 100 ml.
of O·IN acetic acid with o·IN ammonium hydroxide at the
laboratory temperature is shown by the dotted line in Fig.
1-8. The chief feature of the curve is that the change of pH
near the equivalence point and, indeed, during the wliole of
the neutralisation curve is very gradual. The~e is no sudden
change in pH, and hence no sharp end point can be found
with any simple indicator. A mixed indicator, which
* For details. see, e.g .• S. Glasstone. The Electrochemist'l'.Y of Solutions.
p. 224 et seq. (Methuen and Co., 1937.)
exhibits a sharp colour change over a very limited pH range,
may sometimes be found which is suitable. Thus for acetic
acid-ammonia solution tit rations, neutral red-methylene
blue indicator may be used (see .Section 1,29). On the whole,
it is best to avoid titrations involving both a weak acid and a
weak base.
I,36. Neutralisation of a polybasic acid with a strong base.
-The shape of the titration curve will depend upon the
relative magnitudes of the various dissociation constants. It
is assumed that titrations take place at the ordinary labora-
tory temperature in solutions of concentration of O'lN or
stronger. For a dibasic acid, if the difference between the
primary and secqndary dissociation constants is very large
(KlrK'/, > 10,000), the solution behaves like a mixt~re of two
acids with constants Kl and K2 respectively; the considera-
tions given previously may be applied. Thus for sulphurous
acid, Kl = 1·7 X 10- 2 and K2 = 1·0 X 10-7 , it is evident
that there will be a sharp c4ange of pH near the first equiva-
lence point, but for the second stage the change will be less
.pronounced, yet just sufficient for the use of, say, thymol-
phthalein as indicator (see 'Fig. 1-8). For carbonic acid, how-
ever, for which Kl = 4·3 X 10-7 and K'l, = 5·6 X 10-11 ,
only the first stage will be just discernible in the neutralisation
curve (see Fig. 1-8); the second stage is far too weak to exhibit
any point of inflexion and there is no suitable-indicator avail-
able for direct titration. As indicator for the primary stage,
thymol blue may be used (see Section 1,'34), although a mix-
ture of thymol blue (3 parts) and cresol red (1 part) (see Sec-
tion I, 29) is more satisfactory; with phenolphthalein the
colour' change will be somewhat gradual and the titration
error may be several per cent.
It can be shown* that the PH at the first equivalence point
for a dibasic acid is given by
[H+] = JK;K;c-.
+c
Kl
Provided that the first stage of the acid-is weak and that Kl
can be neglected by comparison with c, the concentration of
salt present, this expression reduces to [H+] = vK 1K 2 ,
or pH = 1/2 pK l + 1/2 pK 2 .
With a knowledge of the pH at the stoichiometric point
* See. for ex~mple. S. Glasstone, The Eloctrochemistry 0/ Solutions, 1937,
p.229.
and also of the course of-the neutralisation curve, it should
be an easy matter to select the appropriate indicator for the
titration of any dibasic acid for which Kl/K2 is at least 104 •
For many dibasic acids, however, the two dissociation con-
stants are too close together and it is not possible to differen-
tiate the two stages. If K2 is not less than about 10-7, all
the replaceable hydrogen may be titrated, e.g., sulphuric acid
(primary stage-a strong,acid), oxalic acid, malonic, succinic,
and tartaric acids (for dissociation constants, see Table II,
Section I, 4).
Similar remarks apply to tribasic acids. These may be
illustrated by reference to orthophosphoric acid for which
Kl = 7'5 X 10-3, K2 = 6·2X 10-8, and K3 = 5 X 10-13 • Here
Kl/K2 = 1·2 X lOS and K2/Ka = 1·2 X 105 , so that the
acid will behave as a mixture of three monobasic acids with
the dissociation constants given above. Neutralisation pro-
ceeds almost completely to the end of the primary stage
before the. secondary stage is appreciably affected, and the
secondary stage proceeds almost to completion before the
tertiacy stage is apparent. The PH at the first equivalence
point is given approximately by (1/2 pK1 + 1/2 PK 2) = 4·6,
and at the second equivalence point by (1/2 pK z + 1/2 PK3)
= 9·7; in the very weak third stage, the curve is very flat
and no indicator is available for direct titration. The third
equivalence point may be computed approximately from the
equation (Section I, 19) : -
pH = 1/2 pKw + + 1/2 log c.
1/2 PK"
= 7·0 + 6·15 + (- 0'5).
= 12·6 in O'IN solution.
For the primary stage (phosphoric acid as a monobasic acid),

methyl orange, bromo-cresol green or congo red maybe used
as indicators. The secondary'stage of phosphoric acid is very
weak (see acidK1 = 1 X 10-7 in Fig. 1-8) and the only suitable
simple indicator is thymolphthalein (see Section I, 34); with
phenolphthalein the error may be several per cent. A mixed
indicator composed of pbenolphthalein (2 parts) and a-napli-
tholphthalein (1 part) is the most satisfactory for the deter-
mination of the end point of phosphoric acid as a dibasic
acid (see Section I, 29). The experimental neutralisation
curve of 50 m!. of O·IM orthophosphoric acid with O·IN
potassium hydroxide, determined by electromeric titration,
is shown in Fig. 1-10.
14
13
", ....
12
v-
11
I
10
9
I-~ ~'-~- Phenol phthalein
B
II
pH
7
1/
//
6
6
/
Ir~
~,
4
Methyl orange
3
/
I
2 -
I--'
1:(1 I~ .. Indicator
o Qi :1:.1•
~ :.:~1 Ranges
.
o 10 20 30 40 50 110 10 80 90 190 110 120 130 140 150
Alkali added. mi.
FIG. 1-10. Titration of 50 m1. O'IM H 3 PO, with O'IN KOB.
There are a nu~ber of tribasic acids, e.g., cittic acid with

Kl = 9·2 X 10-4 , K z = 2·7 X 10-5 , Ka = 1·3 X 10--6, the
three dissociation constants 01 which are too close together
and for which the three stages cannot consequently be differ-
entiated. If Ks > ca. 10-7 , all the replaceable hydrogen
may be titrated; the indicator will be determined by the
value of K a•
I, 37. Titration of solutions of hydrolysed salts. Displace-
ment titration.-The examples which are encountered in
practice are largely the salts formed from a strong base and a
very weak ac~d, such as potassium cyanide, borax, and sodium
carbon...te, or the salt~ deriv~d from a strong acid and a very
weak base, for example, aniline hydrochloride. We shall
therefore confine our attention to examples of these salts.
Both types will be largely hydrolyse"d in solution.
Titration of potassium cYaIude with a strong' acid.-Let us
consider first the titration of 100 ml. of O·2N potassium cyan-
ide with 0'2N hydrochloric acid at the laboratory tempera-
ture. 'The salt is completely dissociated:
KeN ~ K+ + CN-.
The hydrolysis of the sa:lt is represented by :
CN- + H 2 0 ~ HCN + OH-.
The reaction upon the addition of hydrochloric acia is simply
·the combination between the H+ ions from the acid and the
OH- ions produced by hydrolysis:
H+' + OH- = H 20.
The net result is that the weak hydrocyanic acid has been
displaced by the stronger hydrochloric acid:
KCN + HCI = HCN + KCI;
the process is therefore often referred to as a displacement
titration. The pH at the true end point can be calculated
from the diss9ciation constant and the concentration of the
weak acid at the equivalence point. Thus in our eXl,lmple,
the volume is doubled at the equivalence point, hence the
concentration ot hydrocyanic acid is O·lN. Ka for hydro-
cyanic acid is 7·2 X "'10-1 °,
i.e., [H+] X [CN-] I [HCN] = 7·2 X 10-1 °;
hence [H+] = v7·2 X 10-10 X 0·1, since [H+l.= [CN-],
or pH = 5·1.
If 100·1 ml. of hydrochloric acid are added, the pH will be
4, with.l00·2 ml. the pH will be 3'7, with 101 m!. the PH will
be 3, etc. There will be a fairly abrupt decrease of pH.in the
neighbourhood of the equivalence point. If the titratjon
error is not to exceed 0·2 per cent, any indicator covering the
pH range 3·7 to_ 5'1 (and slightly beyond this) may be used;
suitable indicators are bromo-cresol gre~n, methyl orange,
bromo-phenol blue,' and methyl red . .
Titration of borax with a strOng acid.-With borax, similar
considerations apply. The net result of the displ_acement
titration is given by': .
Na 2B 4 0 7 + 2HCI + SHiP = 2NaCI + 4H aBO a••
Boric acid behaves as a weak monobasic acid with a dissocia-
tion constant of 6 X 10-1 °. The pH at the equivalence point
. in the titration of 0·2N borax with O·2N hydrochloric acid is
that due to O·lN boric acid, i.e., 5·1. Further p_ddition of
hydrochloric acid will cause a sharp decrease of PH as in the
case of potassium cyanide. Similar indicators may therefore
be used.
84 Quan#tative Inorganic Analysis
Titration .of sodium carbonate with a strOng acid.-The
solution of sodium carboriate may be titrated to the bicar-
bonate stage (i.e., with one equivalent of acid), when the net
reaction is :
Na 2C0 3 + HCI = NaHCOa + NaCl.
The equivalence point for the primary stage of jonisation of
carbonic acid is at pH = (1/2 pK l + 1j2'PK 2 ) = 8·3, and we
have seen (Section I. 34) that .inter alia thymol blue and, less
satisfactorily, phenolphthalein, or a mixed indicator may be
employed to detect the end point.
Sodium carbonate solution may also be titrated until all
the carbonic acid is displaced (two equivalents of acid).
The. net reaction is' then :
Na 2CO a + ~2HCl = 2NaCI + H 2CO a.
The same end point is reached by titrating sodium bicar-
bonate solution with hydrochloric acid:
NaHCO a + HCl = NaCl + H 2C03 •
The end point with 100 ml. of 0'2N sodium bicarbonate and
0·2N hydrochloric acid may be deduced exactly as described
above for potassium cyanide. The pH-at the equivalence
point is very approximately 3·8 (Ki for H 2 CO a = 4·3 X 10-7 ;
the secondary ionisation and the loss of carbonic acid, du~ to
escape of carbon dioxide, has been neglected). Suitable in-
dicators are therefore methyl yellow, methyl orange; congo
red, and bromo-phenol blue. The experimental titration
curve, determined with the hydrogen electrode, for 100 ml.
of O'IN sodium carbonate and O'IN hydrochloric acid is
shown in Fig. l-ll.
Salts of weak bases and strong acids, e.g., aniline hydro-
chloride (Kb = 4·0 X 1O-1()) may be treated in an analogous
manner.
I. 38. Choice of indicators in neutralisation reactions.-As
a general rule it may be stated that fGr a titration to be
feasible, there should be a change of approximately two units
of PH at or near the stoichiometric point produced by the
addition of a small' volume of the reagent. The pH at the
equivalence point may be computed by means of the equations
given in Section I, 19 (see also below); the pH at either siqe .
of the equivalence point (0·1-1 ml.) may be calculated as
described in the preceding sections, and the difference will
indicate whether the change is large enough to permit a sharp
The Tfz,eoretical Basis of Quantitative Analysis 85
13
II ...........
i'- r--.....
~PhenOI
10
............
9
8
'"
j ~PhthaleiO
loymol blue
pH
7 "- .............
6 i'-...
5
i\
4
3
-J~:; orange
I
§ Indicator
Ranges
~
0
0 10 20 30 40 60 60 70 80 90 100
Acid added, ml.
FIG. 1-11. Titration of 1(0 ml. of O'lN Na 2CO a with O'IN Hel.
end point to be determined. Alternatively, the pH change on

both sides of the equivalence point may be 'noted from the
neutralisation curve determined by electromeric titration
(Section 1,52). If th~ pH change is satisfactory, an indicator
should be selecteCl t~at changes at or near the equivalence
point.
For. convenience of reference, we shall summarise the con-
clusions already deduced from theo~etical principles.
Strong acid and strOng base.. For O'IN solutions or stronger,
any iJ)dicator may be used which has a range between the
limits pH 4·5 to pH 9·5. _With O'OIN solutions, the pH
range is somewhat smaller (5'5 to 8·5). If carbon dioxide is
present, the solution should either be boiled whIlst still acid
and the solution titrated when cold, or an indicator with a
range below pH 5 be employed:
Weak acid and a strong base. The pH at the equivalence
point is calculated from the equation:
pH = 1/2 pKw + 1/2 PKa + 1/2 log c.
The PH range for acids with Ka > 10-5 is 7-10'5; for weaker
acids (KII > 10-6 ), the range is reduced (8-10). The pH
range 8-10,5 will cover most of the examples likely to be
encountered; this permits of the use of thymol blue, thymol-
phthalein, or phenolphthalein.
Weak base and strong acid. The pH at the equivalence
point is q>mputed from the equation:
pH = 1/2 pK", - 1/2 PKb - .1/2 log c.
The pH range for bases with Kb > 10-5 is 3-7, and for
weaker bases (Kb > 10-6) 3-5. Suitable i~dicators .will be
methyl red, methyl orange, methyl yell,ow, bromo-cresol
green, and bromo-phenol blue.
Weak acid and weak base. There is no sharp rise in the'
neutralisation curve and, generally, no simple indicator can
be used. The titration should therefore be avoided, if
possible, The approximate pH at the equivalence point can
be computed from the equation:
pH = 1/2 pKfJJ + 1/2 pK" - 1/2 pKb •
It is sometimes possible to employ ~ mixed indicator which
exhibits a colour change over a very limited pH range, for
example, neutral red-methylene blue for ammonia solution
and acetic acid. I
As a general rule, wherever an indicator does not give a

sharp end point, it is advisable to prepare an equal volume
of a comparison solution containing the same quantity of
indicator and of the final products and other components of
the titration as in the solution under test, and to titrate to the
colour shade thus obtained,
THEORY OF PRECIPITATION AND COMPLEX-
FORMAnON PROCESSES
I, 39. Precipitation reactions.- The most important pre-
cipitation processes in volumetric analysis utilise silver
nitrate as the reagent (argentimetric processes). Our dis-
cussion of the theory will therefore be confined to argenti-
metric processes; the same principles can, of course, be
applied to other precipitation reactions. Let us consider the
changes in ionic concentration which occur during the titra-
tion of 100 ml. of O·IN sodium chloride with O'IN silver
nitrate. The solubility product of silver chloride at the
.laboratory temperature is 1·2 X 10-1°.' The initial concen-
tration of chloride ions, [cq, is 0·1 g. equivalent per litre, or
'The Theoreti.cal Basis of Quantitative Analysis 87
pcr = I (see Section I, 17). When 50 mI. of O'IN silver
nitrate have been added, 50 ml. of O'IN sodium chloride
remain in a total volume of 150 ml. : thus [cq = 50 X 0'1/160
= 3,33 X 10-2 , or PCl- = 1'48. With 90 ml. of silver nitrate
solution, [Cr] = 10 X 0'1/190 = 5·3 X 10- 3, or pCl- . 2·28.
Now [Ag+] X [Cr] = 1·2 X 10-10 = SAgCh
or j'JAg+ + pCl- = 9'92.= PAgCl.
In the last-calculation, pCl- = 1'48, hence PAg+ = 9·92
1·48 = 8·44. In this manner, the various concentrations of
chloride and silver ions can be computed up to the equiva-
lence point. At the equivalence point:
Ag+ = Cl- = ~ , •
PAg+ = pcr = 1/2 pAgCI = 9'92/2 = 4·96,
and a saturated solution of silver chloride with no excess of
silver or chloride ions is present.
With -lpO·l mI. of silver :nitrate solution, [Ag+] = 0·1 X'
0'1/200'1 = 5 X 10-5, or pAg+ = 4·30; pCl- = pAgel -
pAg+ = 9·92 - 4·30 = 5'62.*
The values computed in this way up to the addition of llO
ml. of O'IN silver nitrate are collected in Table XIV. Similar
values for the titration of- 100 ml. of O·lN potassium iodide
with O'IN silver nitrate are included in the same Table
(SAgI = 1·7 X 10-16). •
It will be, seen by inspecting the silver ion exponents in the

neighbourhood of the equivalence point (say, between 99·8
and 100·2 ml.) that there is a mar}red change in the silver ion
concentration, although the change is more pronounced for
silver iodide than for the silver chloride since the solubility
producf of the latter is about 1()6 larger than for the former.
This is shown more clearly in the titration curve in Fig. 1-12,
which represents the changes of pAg+ in the range between
10 per cent before and 10 per cent after the stoichiometric
point in the titration of O·IN chloride and O·IN iodide respec-
tively with O'IN silver nitrate. An almost identical curve
is obtained by potentiometric titration with the silver electrode
see (Section 1, 52) ; the PAg+ ~alues may be computed from
the e.m.f. figures exactly as in the calculation of pH.
• This is not strictly true, since the dissolved silver chloride will contribute
silver and chloride ions to the solution; the actual concentration is ca. 1 X 10-'
g. ions per litre.' If the excess of silver ions added is greater than 10 times this
value, i.8., > lO'V'SAgcl, the error introduced by neglecting the ionic con-
centration produced by the dissolved salt may be taken l!s negligible for the
purpose of the ensuing discussion.
TABLE XIV. TITRA"troN 0.1' 100 ML. OF '(HN NaCl AND 100 ML. of O·lN KI
RESPECTIVELY WITH O·IN AgNO, (SAgel = r'2 x 10-10 ; SAgI = 1'7 X
10- 11)
MI. of O·IN Titration of Chloride Titration of iodide

AgNO.
added pCI- PAg+ pr PAg+
0 1·0 - ]·0 , -
• 50 }"·5 8·4 1·5 14'3
90 2·3 7·6 2'3 13'5
95 2·6- 7·3 2·6 13·2
98 3·0 6·9 3·0 12'8
99 3'3 6·6 3·3 12·5
99·5 3·7 6·2 3·7 12·1
99·8 4'0 5·9 4·0 U'S
99'9 4·3 5·6 4'3 11-5
100·0 •
100·1
5·0
5·6
5·0
4·3
7·9
11·5
- 7'9
4'3
100·2 5·9 4'Q ll·8 4·0
100'5 0'3 3·0 12·2 3·6
101 6·6 '3-3 12·5 a·a
102 6·9 3·0 12·8 a·o
105 7'3 2·6 13·2 2·6
110 7·6 2'3 13·5 2·4
14
~
f--
13 r- Agi
i\
11
10
.
s
.
8
pA + r-- ~<f!-t.CJ~ r~-p:SAgI
9 7
8 r'\
6 pSAgCI
,
4
3 1"- ~~~g}-
AgCJ+:A;+-
90 92 94 96 98 100 102 104 106 IDS 110

0·' N AgN03 added. mI.
FIG. 1-12. Titration Curv~s of 100 ml. of O·IN·NaCl and -of 100 ml. of O'lN KI
respectively with O·lN AgNO. (Calculated).
..
I, 40. Complex formation reactions.-An example of this

type of reaction is the titration of cyanide with silver nitrate
solutions, a method first 'proposed by Liebig (1851). When a
solution of ,silver nitrate is' added to a solution containing
cyanide ions (e.g., an alkali cyanide) a white precipitate is
formed when the two liquids first come into contact with one
anoth~r, but on stirring it redissolves owing to the formation
of a stable complex cyanide, the alkali salt of which is soluble:
AgN0 3 + 2KCN ~ K[Ag(CN)2J + KNO a,
or Ag+ + 2CN- ~ [Ag(CN)J-:
When the above reaction is complete, further addition of
silver nitrate solution yields the insoluble silver argento-
cyanide (often termed silver cyanide) :
Ag+ + [Ag(CN)J- ~ Ag[Ag(CN)2J.
The end point of the reaction is therefore indicated by the
formation of a permanent precipitate or turbidity.
Let us attempt to investigate the suitability of this.reaction
for quantitative work and to calculate the titration error
involved. The instability constant (Section I, 12; Table V),
i.e.,
[Ag+] X [CN-];2 j [f Ag(CN)2 r],
is 1·0 X 10- at the ordinary temperature. Let us suppose,
21
for purposes of illustration, that an amount of potassium

cyanide solution equivalent to 10 ml. of O'IN silver nitrate
is employed, and that the volume at th~ end point is 100 ml.
The concentration of the complex K[Ag(CN)2] will then be
0·01 molar. Here the Ag+ and CN- ions present will be
derived from the dissociation of the complex [Ag(CN)z]-
ion. At the stoichiometric point:
[CN-] = 2[Ag+J = 2x,
hence [Ag+] X [CN-P = 1.0 10- 21 = % X (2%) z

[{ Ag(CNh} ] X 0'01'
or x = V2:5 X 10-3 = 1·36.x 10- 8,
or [Ag+] = 1·36 X 10-8 , and [CN-J = 2·72 X 10-8 .
The solubility product of silver argentocyanide, [Ag+j X
[{ Ag(CN)z rJ, is 2·25 x-JO-12, hence precipitation of this
substance will commence when:
+ 2·25 X 10-12 2·25 X 10-12
[Ag J > [{Ag(CNhrl > 0.01 > 2·25 X 10-1°,
The silver ion concentration at the theoretical end,. point is
1·36 X 10-8, hence precipitation will occur just before this.
The difference is so small, however, that for all practical
purposes the titration error is negligible; moreover, in prac-
tice, the theoretical end point will be exceeded in order to
produce enough predpitate to be visible.
The above method yields satisfactory results if the reagent
is added very slowly near the end point, but is not applicable
to ammoniacal solutions owing to the removal of the silver
ions by the formation of the soluble complex [Ag(NHa)2]+
(see Section I, 12). An alternative procedure is to work in
ammoniacal solution in the presence of a little alkali iodide
(G. Dcniges, .1895) : the end point is marked by the formation
of a turbidity of silver iodide (see Sections I, 11, 12).
I, 41. Determination of end pointa in precipitatioh and in
complex formation reactions.-Many methods are utilised in
determining end points in these reactions, but only the most
important will be mentioned here.
A. Formation of a coloured precipitate. This maybe
illustrated by the Mohr procedure (1856) fC?r the determina-
tion of chloride and brorpide. In the titration of a neutral
solution of, ~ay, choride ions with silver nitrate solution, a
small quantity of potassium chromate solution is added to
serve as indicator. At the end point the .chromate ions
combine with silver ions to form the sparingly soluble, red
silver chromate.
The theory of the process is as follows. We have here a case
of fractional precipitation (Section I, 11), the two sparingly
soluble salts being silver chloride (S.P. 1·2 X 10-1 °) and silver
chromate (S.P. 1·7 X 10-12). Let us consider an actual
example encounteted in practice, viz., the titration of, say,
O'IN silver nitrate with O'IN sodium chloride in tne presence'
of a few m!. of dilute potassium chromate solution. Silver
chloride is the less soluble salt and, furthermore, ·the initial
chloride concentratiQIl is high, hence silver chloride will be
precipitated. At the first'point where red silver chromate is
just precipitated, we shall have both salts in equilibrium with
the solution, hence :
[Ag+] X [C1-] = SAgCl = 1·2 X 10-10 ;
[Ag+12 X [Cr0 4-j ,_ SAg2CrOc = 1·7 X 10-12 ;
SAgel
[Ag+]
[Q-] =
•
[cq _
SAgCl
-
1·2 X 10- 10
= 9·2 X 10- 5
v[crO, 1 v'SAg2CrQ' v1'7 X 10-12
At the equivalence point [cq = ~ = 1·1 X 10-5 •
If silver chromate is to precipitate at this chloride ion con-
centration :
[CrO --1 = ( [Cn )2
= (1.1 X 10- )2=1.4 X 10-B
5
, 9·2 X 10-6 9·2 X 10-6
or the potassium chromate solution should be 0·014 molar.
It should be noted that a slight excess of silver nitrate solution
must be a{jded \ before the red colour of silver chromate is
visible. In practice, a more dilute solution (0'002-0'004
molar) of potassium chromate is generally used, since a
chromate solution of concentration 0·01-0·02 molar imparts
a distinct deep orange colour ·to the solution, which renders
the detection of the first appearance of' silver chromate
somewhat difficult. We can readily calculate the error there-
by introduced, for if [CrO,--] = (say) 0·003, silver chrorpate
will be precipitated when:
[Ag+] =
'\I~
~ =J1.7 X 10-
3 X 10- 3
12
= 2·4 X 10--5.
If the theoretical concentration of indicator is used :
[Ag+]
•
=
J l'7 X 10-12
1·4 X 10- 2
= 1·1 X 10-5 •
The difference is 1·3 X 10--5 g. equivalent per litre. If the

volume of the'solution at the equivalent point is 150 ml.,
then this corresponds to 1·3 X 10-5 X .150 X 10' I 1000 =
0·02 ml. of O·lN silver nitrate. This is the theoretical titra-
tion error, and is therefore negligible. In actual practice
another factor must be considered, viz., the small excess_of
silver nitrate solution which must be added before the eye can
detect the colour change in the solution; this is of the order
of one drop or ca, 0·05 mI. of O·IN. silver nitrate.
The titration error will increase with increasing dilution of
the solution being titrated and is quite appreciable (ca. 0·6
per cent) in dilute, say, O'OlN, solutions when the chromate
concentration is of the order 0'002-0·004 molar. This is
most simply allowed for by means of an indicator blank
determination, e.g., by measuring the volume of standard
silver nitrate solution required to give a perceptible .coloura-
tion when added to distilled water containing the same
92 Quani#ative Inorganic Analysis~
quantity of indicator as is employed in the titration. This
volume is subtracted from the volume Qf standard solution
used.
It must be mentioned that the titration should be carried
out in neutral solution or in very faintly alkaline solution,
i.e., witilin the pH range 6·5 to 9. In acid solution, the
following reaction occurs:
2Cr0 4- - + 2H+ ~ 2HCr0 4- ~ Cr 20 7- - + H 20.
HCr0 4- is a weak acid, consequently the chromate ion con-
centration is reduced and the sol).lbility product of silver
chromate may- not be exceeded. In markedly alkaline
solutiqns, silver p.ydroxide (S.P. 2·3 X 10- 8) might be pre-
cipitated. A simple method of m.aking an acid solution neu-
tral is to add an excess of pure calcium carbonate.
B. Formation of a soluble coloured compound. This
procedure is exemplified by the method of Volhard (1878)
for the titration of silver in the presence of free nitric acid
with standard potassium or ammonium thiocyanate solution.
The indicator is a solution of ferric nitrate or of 'ferric am-
monium alum. The addition of the thiocyanate solution
produces first a precipitate of silver thiocyanate (S.P. 7·1 X
10-13) :
Ag+ + CNS- ~ AgCNS.
When this reaction is cOIl}.plete, the slightest excess of
thiocyanate produces a reddish-brown colouration, due to the
formation ,9£ 'the complex fern-thiocyanate ion. * •
Fe+++ + 6CNS- ~ [Fe(CNS)6r --
(or 2Fe+++ + 6CNS- ~ Fe[Fe(CNS)6J
~ Fe+++ + [Fe(CNS)6r--).
This method may be applied to the determination of

chlorides, bromides and iodides in acid solution. Excess of
standard silver nitrate solution is added and the excess is
back-titrated with standard thiocyanate solution. For the
chloride estiination, we have the following two equilibria
during the titration of excess of silver ions:
Ag+ + Cl- ~ AgCl;
Ag+ + CNS- ~ AgCNS.
The two sparingl~' soluple salts 'will be in equilibrium with
the solution, hence:
[Cn SAgCl 1·2 X 10-10
= 170.
[CNS J SAgCNS 7·1 X 10-13
... According to Bent and French [1941] the colouration is due to the
FeCNS++ion:
Fe + + + +CNS- ~ FeCNS+ +.
£

When the excess of silver has reacted, the excess of
thiocyanate may react with the silver chloride, since
silver thiocyan~te is the less soluble salt until the ratio
[cq I[CNS-] in the solution is 170: .
AgCl + CNS- ~ AgCNS + CC
This will take 'place before reaction occurs with the ferric ions
in the solution, and there will consequently be a considerable
titration error. It is therefore. absolutely necessary to pre-
vent the reaction between the thiocyanate and the silver
chloride. This ·may be effected in several ways, of which the
first is probably the most reliable: (i) the silver chloride is
filtered off, and the filtrate and washings are titrated;
(ii) the silver chloride is coagulated, and thus rendered less
reactive by boiling, and the mixture is cooled and titrated;
(iii) the addition of a few ml. of ether or, preferably, of 1 ml.
of nitroben~ene, which causes coagulation, before back
titration.
Wit~ bromides, we have the'equilibrium :
[Br-] SAgBr 3·5 X 10-13
0·5.
[CNS-] = SAgCNS - 7·1 X 10-13 -
The titration error is small and no difficulties arise in the
determination of the end point. Silver iodide (S.P. 1·7 X
10-16 ) is less soluble than the bromide·; the titration error is
negligible, but the ferric ir~m indicator should not be added
until excess or silver is present since the .dissolved iodide
reacts with the ferric iron:
2Fe+++ + 21- ~ 2Fe++ :t- J2'
C. Use of adsorption indicators. K. Fajans (1923-1924) has
introduced a new type of indicator for pr~cipitation reactions
as a result of his studies on the nature of adsorption. The
action of these indicators is due to the fact th'l-t at the equiva-
lence point the indicator is adsorb~d by the precipitate, and
during the process of adsorption a change occurs in the
indicator which leads to a substance of different colour;
they have therefore been termed adsorption indicators. The
substances employed are either acid dyes such as those of the
fluorescein series, e.g., fluorescein and eosin which are utilised
as the sodium salts, or basic dyes, such as those of the rhoda-
mine series (e.g., rhodamine 6G), which are applied as the
halogen salts.
The theory of the action of these indicators is based upon
the properties of colloids (see Section I, 56). When silver
nitrate solution is titrated with a solution of a chloride, the
precipitated silver chlotide adsorbs chloride ions (a precipitate
has a tendency to adsorb its own ions); this may be termed
the primarily adsorbed layer, and it will hold by secondary
adsorption oppositely charged ion present in solution (shown
diagrammatically in Fig. I-I3a). As soon as. the stoichio-
me_tric point is reached, silver ions are pre~ent in excess;
these will now be primarily adsorbed and nitrate ions will be
held by secondary adsorption (Fig. 1-13 b)_ If fluorescein
is also present in the solution, the negative "fluorescein ion,
which is much more strongly adsorbed than the nitrate ion, is
immediately adsorbed, and will reveal its presence in the
precipitate, not by its own colour, which is that of the solution,
but by the. formation of the red (or pi'nk) silver fluoresceinate
on the surface with first trace of excess of silver ions.
:8\..
NO;
, --~g-":"" ,
,, .,
,,
NO.1 :Ag+ Agel Ag+;NOj
.
,, ,'
. ,,
"
...... _- ... -_ .... "
I
,
(a) . ( b)
FIG. 1-13. (a) 4gCl precipitated in the presence of excess of CC
(b) Agel precipitated in the presence of excess of Ag+.
An alternative view is that during the adsorption of the

fluorescein ion a re-arrangement of the structure of the ion
occurs with the formation of a coloured substance. It is
important to notici that the colour change takes place at the
surface of the precipitate. If chloride is now added, the
suspension will remain p~nk until chloride ions are present in
excess, the adsorbed silver will then be converted into silver
chloride, which will then primarily adsorb chloride ions. The
fluorescein ions secondarily adsorbed will pass back into
solution, to which they impart a greenish-yellow colour_
The following conditions will govern the choice of a
suitable indicator:
(i) The precipitate should separate as far as possible in the
colloidal condition. Large quantities of neutral saIts,
The Tlieoretical 'Basis of Quantitati'lJ.t Analysis 95
particularly of polyvalent ions, should be avoided owing to
their coagulating effect. The solution should not be too
dilute as the amount of precipitate formed will be small and
the colour change far from sharp with certain indicators .•
(ii) The indicator ion must be of, opposite charge to the ion
of the precipitating agent.
(iii) The indicator ion should not be adsorbed before the
particular compound has been completely precipitated, but it
should be strongly adsorbed immediately after the equivalence
point. The indicator ion should not be too strongly adsorbed
by the precipitate; -if this occurs, the adsorption of the indicator
ion may be a primary process and will take place befor~
the equivalence point, e.g., eosin (tetrabromofIubresc~in) in
the chloride-silver titration. •
For the titration of chlorides, fluorescein may be used.
This indicator is a weak acid (Ka = ca. 1 X 10-8) and cannot
be employed in strongly acid solutions. Titration should be
carried out in neutral or in feebly acid (acetic acid) solutions;
acid solutions may be treated with a slight excess of sodium
acetate. Alternatively, dichlorofluorescein, which is a
stronger acid, may be utilised in slightly acid solutions
(PH> 4·4); this indicator has a further advantage in th<tt it
is applicable in very dilute solutions. -
. For the titration of bromides, iodides and thiocyanates
with silver, eosin is a suitable indicator; the coloul' on the
precipitate is magenta. The best results for iodides a~e
obtain,ed with dimethyldi-iodofluorescein, which gives a blue-
red colour on the precipitate. The titration of silver solutions
with bromide as the precipitating agent is best effected with
rhodamine 6G, the precipitate assuming a violet colour at the
end point; a sharp end point is obtained with this indicator
in the presence of nitric acid up to a concentration of 2N.
Cyanides may be titrated.. with standard silver solutions using
diphenylcarbazide as the adsorption indicator (see Section
I, 41); the precipitate is pale-violet at the end point.
D. Turbidity method. The appearance of a turbidity is
sometimes utilised to mark the end poi.nt of a reaction, as in
Liebig's method for cyanides (see Section I, 41). A method
which should be included here is the equal turbidity procedure
introduced by Gay Lussac (1832) for the determination of
silver with chloride. A.t the end point of the titration of,
say, sodium chloride with silver nitrate solution, a solution of
sodium nitrate which is saturatea with silver chloride is
obtained. If the clear supernatant liquid is withdrawn, and
equal volumes treated with equivalent strengths of silver
nitrate and sodium chloride solutions respectively, then equal
turQidities of silver chloride should be produced. If equal
turbidities are not found, the titration is continued with
sodium chloride or silver nitrate to the .point of equal
turbidities. .
This method yields very accurate results in experienced
hands, and is still employed for the determination of silver in
alloys in the mints of certain governments. For ordinary
work, however, any of the methods described above will be
fOUIid more suitable for- the average student. It must,
however, be pointed out that nephelometric methods are
employed in atomic weight determiI1ations ~nd, indeed, owe
their introduction to T. W. Richards who first used a simple,
but -effective, nephelometer in connexion with his researches
on the atomic weight of silver (compare Section V, 5).
I, 42. THEORY qF OXIDATION-REDUCTION

REACTIONS
Reference has already been made in Section I, 23 to the
fact that oxidation may be defined as the loss of electrons, and
reduction as the gain of electrons by an atom or group of
combined atoms. Thus in the reduction of ferric chloride
by stannous chloride:
2FeCIa + SnC1 2 +
or 2Fe+++ + Sn++ 2Fe++ + Sn++++ ;
for every gram atom of ferric iron (56 g.) reduced 96,500
coulombs or one Faraday of electricity is lost by the iron, and
for every gram atom of stannous tin (119 g.) oxidised, the
latter gains 2 X 96,500 coulombs or two Faradays.*
According to modern theory, an el~cfric current is essen-
tially a transfer of electrons. It should theref_ore be possible
to obtain direct proof of the transfer of electricity in the
oxidation-reduction reaction under suitable experimental
oonditions. This may be shown by the following experiment
• It follows from Faraday's laws (Section It 64) that each gram equivalent
weight of an ion is associated with a charge of one Faraday of electricity. A
change of charge of one thus corresponds to the gain or loss of one Faraday
per formula weight of the substance.
The Theoretical Basis of Quantitative Analyst's 97
(Fig. 1-14). Solutions of stannous ,chloride and of ferric
chloride, each acidified with dilute hydrochloric acid in order
to increase the conductivity, are placed in separate beakers A
and B, and the two
solutions are connected v
by means of a "salt
b rid g e" containing
sodi~m chloride. The
latter consists of a
t 1
U-tube filled with a
solution of a conducting NaCI
eledrolyte, s u c has
sodium chloride, and
stoppered at each end
with a plug of cotton
wool in order to min-
imise diffusion. It con- A B I
nects the two parts of , FIG, 1-14.
the redox system, and
prevents mixing. The electrolyte in solution in the salt bridge is
always selected so that it does not react chemically with either
of the solutions which it conne,:ts. Platinum foil electrodes
are introduced i{J,to each of the solutions, and the two elec-
trodes are connected to.a milli-voltmeter V. When the cir-
cuit is closed, it will be found that the current in the external
circuit passes from the stannous chloride solution to the ferric
chloride solution. After a time, stannic ions can be detected
in A and ferrous ions 'in B.
In the preceding experiment the flow of current is due to
the e.m.f. produced by the chemical change in the redox
reaction. It can also be shown that an electric current, say
from a storage battery, will give rise to oxidatiop.-reduction
reactions. Thus, if in Fig. 1-14 beaker A contains ferrous
ammonium sulphate acidified with dilute sulphuric acid, and
beaker B ferric chloride likewise acidified with dilute sulphuric
acid, the salt bridge dilute -sulphuric acid, and the milli-
voltmeterV replaced by a 6-volt battery, then after sOlI"le time
ferric ions can be detected in A and ferrous ions in B.
It is a well-known fact that oxidants and reductants vary
considerably i1l strength. Information of a qualitative
character as to the relati.ve strengths (compare relative
strengths of weak acids and bases as expressed by .th~ir dis-
sociation constants in Table II) would be of considerable
practical value. A method which suggests itself at once is
E
the measurement of the e.m.f. of redox reactions under
standard conditions. This is indeed the method which is
employed, but before this can be understood by the student,
it will be necessary to consider the subject of electrode poten-
"Hals.
I, 43. Electrode potentials.-When a metal is immersed
in a solution containing its own ions, say, zinc in zinc sulphate
solution, a potential difference E is established between the
metal and the solution. The potential difference E for ad
electrode reaction, M ~Mn+ +n e, is given by the expression* :
• RT 1oge a+
E = nF +E 0
0'1'),
"
where R is the gas constant, T is the absolute temperature,

F the Faraday, n the valency of the ions, a+. the activity of
the ions in the solution, and Eo is a constant dependent upon
the metal. Equation (i) can be simplified by introducing the
known values of Rand F, and converting the natural loga-
rithms to base 10 by multiplying by 2·3026; i~ then Becomes:
-_ {j'0001982T 1 E (..
E
n
og a+ + " 11).
For a temperature of 25°C (T = 288°) :
0'0591 I
E = - - og a+ + Eo (iii) .
n
For most purposes in quantitative analysis, it ~s sufficiently
accurate to substitute a+ by c,+, the ion concentration (in
mols per litre) :
0·0591
E = - - log C,+
n
+ Eo (iv).
The latter is often known as the Nernst equation.
For a non-metal, which yields negative ions, equation (iii)
becomes:
0·0591
E = Eo - - _ log a_ (v),
n
where ~_ is the activity of the negative ions. Here also, we
may introduce the approximation of substituting Ci- for a_.
If in equation (iv) , c;+ is put equal to unity, E is equal to
Eo. Eo is called the standard or normal potential of the metal.
• For a aerivation of this equation, see any good text book of physical
chemistry, or the author's Text Book of Qualitative Chemical AnalYSis, 1937,
p.67. (Longmans Green and Co. Ltd.).
In order to determine the potential dif£<rrence between an
electrode and a solutiun, it is necessary to have another
electrode and a solution, the potential difference between
which is known. The two electrodes can then be combined
to form a voltaic cell, the e.mJ. of which can be directly
measured. The e.mJ. of the cell is the arithmetical sum or
diffetence of the electrode potentials (depending upon the
sign of these two potentials); the value of the unknown
potential can then be calculated. In practice the standard
electrode potential used for comparative purposes is the
mow or normal hydrogen elec- 0
trode. This consists of a piece
of platinum foil, coated with
platinum black by an electro-
lytic process, and immersed in
a solution of hydrochloric acid, A
normal with respect to hydro-
gen ions (more correctly, .in a
solution of hydrochloric acid
containing hydrogen ions of
unit activity). Hydrogen gas •
at a pressure of one atmosphere
is passed over the foil through
the side tUDe C (Fig. 1-15) and
escapes through the small
openings B in the surrounding FIG. 1·15.
glass tube A; the foil is thus kept saturated with .the gas.
Connection between the platinum foil sealed into the tube D
with an outer circuit is made with mercury in D. The plati-
niun black has the remarkable property of adsorbing large
quantities of hydrogen, and it permits the change from the
gaseous to the ionic form and the reverse process to occur
without hindrance; it therefore behaves as though it were
composed entirely of hydrogen, that is, as a hydrogen elec-
trode. Under fixed conditions, viz., of atmospheric pressure
and normal concentration of hydr<?gen ions in the solution in
contact with the electrode, the hydrogen electrode possesses
a definite potential at all temperatures. By connecting the
normal hydrogen electrode with a metal electrode (a metal
in contact with a molar solution of its ions) by means of a
salt (say, potassium chloride) bridge, the normal, molar.
or standard electrode potential may be determined. Other elec-
trodes, particularly the calomel electrode and the silver-silver
chloride electrode, the potentials of which oha ve been determined
100 Quantitative I norganic ~ nalysis
by direct reference to the normal hydrogen electr04e,

are often used in practice owing to convenience of manipula-
tion. The former is discussed in Section I, 52, but fo.r a
fuller treatment the student is referred to text books of electro-
chemistry. *
We may therefore define the standard or normal electrode
potential of an element as the e.m.f. produced when a half cell
consisting of the *element immersed in a molar solution of its
ions (more correctly, in a solution of its ions possessing unit
activity) is coupled with a molar 'hydrogen electrode, the
potential of which is assumed equal to zero. The more
important standard electrode potentials at 25°C are collected
in Table XV; the sign of the potential here adopted is that
of the charge on the electrode.
TABLE XV., STANDARD ELECTRODE PO~ENTIALS AT 25° c.
LiJLi+ - 2·959 c.ojCo++ -0·28

KjK+ - 2·924 NijNi++ -0·23
Ca/Ca+_+ -2·76 Sn/Sn++ - 0:136 ~
NajNa+ - 2,715 • PbjPb++ - 0·122
MgjMg++ -1·55, Pt(H;) flj: + 0·000
AljAl+++ -1·33 BijBi+++ + 0·226
Mn/Mn++ - 1·1 Cu/Cu++ +. 0·344
ZnjZn++ - 0·762 HgjHg2++ + 0·799
Cr/Cr++ - 0'557 AgjAg+ + 0·798
Fe/Fe++ - 0·441 PdjPd++ + 0:82
Cd/~d++ - (HOI Au/Au+++ + 1·36
When metals are arranged in the order of their standa-rd
electrode potentials, the so-called electrochemical series of
the metals is obtained. The greater the negative value of
the PQtential, the greater is the tendency of the metal to pass
into the ionic state. A metal with a. more negative potential
will displace any other metal below it in the series from
solutions of its salts. Thus magnesium, aluminium, zinc or
iron will displace copper from soluti\)ns of its salts; 'lead will
displace copper, mercury or silver; copper will displace silver.
The standard electrode potential is a quantitative measure
of the readiness of the element to lose electrons. I t is there-
fore a measure of the strength of the element as a reducing
agent; the more negative the element. the more powerful is
its action as a reductant.
• See, for example, S. Glasstone, The Electrochemistry of Solutions (1937).
The: Theoretical Basis of Quantitatlve Analysis 101
1.44. Concentration cells.-The electrode potential varies
with the concentration of the ions in the solution. Hence by
bringing two electrodes of the same metal, but immersed in
solutions containing different concentrations of the ions into
contact, a cell may be formed. Such a cell is termed a
concentration cell. The e:m.f. of the cell will be the algebraic
difference of the two potentials, if a salt bridge be inserted
to eliminate the liquid-liquid junction potential. It may be
calculated as follows. At 25°C:
0·0591 0·0591
E =
n
log C1 + Eo - -n- log C2 + Eo,
0'0591 c
log - 1 , where
1 C > l).
n· Cz
As an example we may consider the cell :

+ .-
Ag r AgNO a aq. II
AgNO a aq.
[Ag+] = O'00475M [.~g+] = 0'043M
I A
g.
E2 E1
Assuming that there is' no potential difference at the liquid
junction:
0·0;591 0·043
E = El - E2 = - - log = 0·056 volt.
1 0·00475
I, 45. Calculation of the e.m.f. of a voltaic cell.-An
interesting application of electrode potentials is to the calcu-
la,tion of the e.m.f. of a voltaic cell. One of ~he simplest of
galvanic cells is the Daniell cell. It consists of a rod of zinc
dipping into 'zinc sulphate solution and a strip of copper in
copper sulphate solution; the two solutions are generally
separated by placing one inside a porous pot and the other
_in the surrounding vessel. The cell may be represented as :
Zn'l ZnSO, aq.11 CuSO, aq.1 Cu.
At the zinc electrode, zinc ions pass into solution leaving an
equivalent negative charge on the metal. Copper ions are
deposited at the . copper electrode, rendering it positively
charged. By completing the external circuit, the current
(electrons) passes from the :mnc to the copper. The chemical
reactions in the cell are as follows:
." (a) zinc electrode- Zn ~ Zn++ + 2E;
(b) copper electrode- Cu++ ~ Cu - 2E.
102 .Quantitative Inorgam'c Analyst's
The net chemical reaction , is ;
Zn + Cu++ = ZnH + Cu.
The potentral difference at each electrode may be calculated
by the formula given above, and the e.m.f. of the cell is the
algebraic difference of the two potentials, the correct sign
being applied to each.
As an example we may ca1culate the e.m.f. of the Daniell
cell with molar concentrations of zinc ions and cupric ions.
E = Eo(cu) - EO(Zn) = + 0·34 - (-0·76) = 1·10 volts.
The small potential difference produced at the contact be-

tween the two solutions (the so-called liquid-junction poten-
tial) is neglected. ,
1, 46. Oxidation reduction cells.-Reduction is accompanied
by a gain of electrons and oxidation by a loss of electrons.
In a system containing both an oxidising agent and its reduc-
tion product, there will be an equilibrium between them 'and
electrons. If an inert electrode, such as platinum, is placed
in a redox system, for example one containing ferric and
ferrous ions, it will assume a definite potential indicative of
the position of equilibrium. If the gxidation tendency pre-
dominates, the system will take electrons from the platinum,
leaving the latter positively charged; if, however, the system
has reducing properties, electrons will be given up to the-
metal, which will then acquire a negative charge. The
magnitude of the potential will thus be a measure of the
oxidising or reducing properties of the system,
To obtain comparative values of the "strengths" of
oxidising agents, it is necessary, as in the case of the electrode
potentials of the metals, to measure under standard experi-
mental conditions the 'potential difference between the
platinum and the solution relative to a standard of reference.
The primary standard is the molar or normal hydrogen
electrode (Section I,' 43), and its potential is taken as zero:
The standard experimental conditions for the redox system
are those in which the ratio of the concentration of the oxidant
to that 'Of the reductant is unity. Thus .for a ferrous-ferric
chloride electrode, the redox cell would be:
Pt \ ~:::+ I HT 1 Hz (Pt).
The potential measured in this way is called the standaid
oxidation potential. A selectio~ of these is given in Table
)fVI. The sign of the potential is that of the electrode.
The Theoretical, Basis of Quantitative Analysis 103
TABLE XVI. STANDARD OXIDATION POTENTIALS AT 25°C.
E'
Electrode Electrode Reaction (volts)
Co+++, Co++ I Pt· Co+++ + ( ;=' Co++ + 1-82
Pb++++, Pb++ I Pt Pb-l;+++ + 21 ;=' Pb++ + }'75
MnO,-, Mn++ I Pt MnO,- + 8H+ + 51 ;=' Mn++ + 4HaO + 1·52
Ce++++, Ce+++ / Pt Ce+++++ l ;=' Ce+++ + 1·45
CIO.-, Cl- I Pt CIO.- + 6H+ + 6. ~ CI- + 3H,O + 1-45
Br0 3-, Br- I Pt BrO.- + 6H+ + 6. ~ Br- + 3H,O + 1·42
CI., 2CI- IPt + 2( ;=' 2CI- + 1·:>6
Crp,--, 2Cr+++ I Pt Crp,-- + ~H+ + 6( ~ 2Cr+++ + 7HP + 1·3
10 3-, r I Pt 10.- + 6H+ + S( ;=' r- + 3H,0 + 1·2
Br,. 2Br- I Pt Br. + 2( ;=' l3r- + 1·07
Fe+++, Fe++ I Pt Fe+++ + ( ~ Fe++ , + 0·75
H.AsO" H,AsO.1 Pt Ha AsO, + 2H+ + 2. ~ H.AsO, + H,O + 0·57
I., 2r I Pt I. + 2! ;=' 21- + 0·54
Fe(CN).---
Fe(CN).---- IPt Fe(CN)a--- + f~ Fe(CN)a---- + 0·49
UO,++. u++++ / Pt UO.++ + 4H+ + 21 ;=' U++++ + 2H.O + 0·36
Cu++, Gu+ / Pt Cu++ + ;;;:Cu+ € + 0·16
Sn++++. Sn++ / Pt Sn++++ + 2( ~ Sn++ + 0-14
Ti++++. Ti+++ I Pt Ti++++ + ~ Ti+++
f + 0·04
H.,2H+ IPt 2H+ + '2( ;=' H, 0·00
Cr+++, Cr++ I Pt Cr+++ + ( ~ Cr++ - '0'4
So, S-- I Pt So + 21 ;=' S-- - 0·55
The oxidation potentials enable us to predict which ions

will oxidise or reduce other ions at molar concentrations.
The most powerful oxidising agents are those at the upper
end of the Table, and the most powerful reducing agents at
the lower end. Thus permanganate ions can oxidise Cl-,
Br-, C, Fe++ and Fe(CN)6- - - -; ferrle ions can oxidise
HaAsOa and I-, but not Cr 2 0 7- - or Cl- ions. It must be
emphasised that oxidation potentials do not give any in-
formation as to the speed of the reaction; in some cases, as in
the oxidation of arsenious acid by eerie ions, the presence of a
catalyst 'is necessary in order that the reaction ll}ay proceed
with reasonable velocity.
I, 47. Calculation of the oxidation'potential.-A reversible
oxidation-reduction reaction may be written in the form
(oxidant = substance in oxidised state, reductant = substance
in reduced state) :
Reductant ~ Oxidant n E. +
'--y-___.) '-----v---J
Red Ox
• The symbol Pt throughout this Table must be interpreted as any inel:',t,
unattacbble electrode,
The electrode potential which is established when an inert or
unattackable electrode is immersed in a solution containing
both oxidant and reductant is given by the expression:
ET = Eo + RTI tlo"
-,- og.- (i) ,
nP aRed
where ET is the observed potential of the redox electrode at
r, EO is the standard oxidising potential, n the number of
electrons gained by the oxidant in. being converted into the
reductant, and aox and aRed are the activities of the oxidant
and reductant respectively.
Since activities are often difficult to determine directly,
they may be replaced by concentrations; the error thereby
introduced is usually of no great importance. The equation
therefore becomes:
RT I
E T = E o'+-P Cox
• (i ').
oge~
n CRed
Substituting the known values of Rand P, and changing
from natural to common logarithms (by multiplying by
2'3026), we have for a temperature of 25°C (T = 288 0
) :
E - EO 0,0591 I [Ox]
Zoo - + - n - og [Red] (i ").
If the concentrations (or, more accurately, the activities) of

the oxidant and reductant are equal, E 20' = EO, i.e., the
standard oxidation potential. It follows from this exptession
that, for example, a ten-fold change in the ratio of the con-
centrations of the oxidant to the reductant will produce a
change in the potential of the system, of O'0591/n volts.
I, 48. Equilibrium constants of oxidation-reduction reac-

tions.-The general equation for an oxidation-reduction
electrode may be written:
pA + qB + rC + ..... N€ ~ sX + ty + uZ + .....
The potential is given by :
RT ax . a~ . a~ ..... .
E = EO + -P
n
loge p q r'
a A • a B • Cfc . .... .
'
.
where a reters to activities, and n to the number of electrons'
The Theoretical Basis oj Quantitative Analysfs 105
involved in the oxidation-reduction reaction. This expression
reduces to the following fop a temperature of 25°C (.concen.tra-
,tions are substituted for activities in order to permit of its
facile application in practice) :
E = EO 0·0591 c~ . c~ . c~ ..... .
+ - - l o g 1> q r
n CA' CB • Cc • • • • • •
It is, of course, possible to calculate the influence of the change

01 concentration of certain constituents of the system by the
u..se of the latter equation. Consider, for example, the per-
manganate reaction:
MnOi- + 8H+ +.5E ~ Mn++ + 4H 2 0.
0·0591 [MnO -:-] X [H+]8*
E = EO + - - log " (at 25°C).
5 [Mn++]
The concentration (or activity) of the water is taken as con-
stant, since it is assumed that the reaction takes place in
dilute solution, and the concentration of the water does not
change appreciably as the result of the reaction. The equa-
tion may be written in the form:
E = EO 0'0591 [MnO,-] 0'0591 [H+]8.
+ 5 1og [Mn++] + 5 10g
This enables us to calculate the effect of change in the ratio
[MnO,-] / [Mn++] at constant hydrogen Ion concentration,
other factors being maintained constant. In this particular
case-difficulties are experienced in the latter calculation owing
to the fact that 'he reduction products of the permanganate
ion vary at different hydrogen ion concentrations. In other
cases no such difficulties arise, and the calculation may be'
employed with confidence. Thus in the reaction:
HsAsO" + 2H+ + 2e ?HaAsOa + H 20,

E = EO + 0'.0~11og [~sAs6,,] X [H+J2 (at 250C),
2 [HsAs03]
_ EO 0'0591 [HsAsO,l 0'0591 1 [H+]Z
or E - + 2 1og [H3AsOSJ + 2 og .
We are now in a position to calculate the equilibrium con-
stants of oxidation-reduction reactions, and thus to determine
• More accurately, activities should be employed. The square brackets
denote molecular concentrations.
E*
106 Qu~ntitative Inorganic ;Analysis
whether such reactions can find application in quantitative
anaJysis. Let us consider first the simple reaction:
C1 2 + 2Fe++ ~ 2Cl- + 2Fe+++.
The equilibrium constant is given by :
[Cq2 x;[Fe+++]2 K.
[C1 2] X [Fe++p ;:-
The reaction may be regarded as taking place in a voltai.c
cell, the two half ~ells being a CI 2 ,2Cl- electrode and a Fe++,
Fe+++ electrode. The reaction is allowed to proceed to
equilibrium, tht; total voltage or e.m.f. of the cell will then be
zero, i.e., the.potentials of the hyo electrodes will be equal.
0·059 [C1 2] 0·059 [Fe+++]
EOcll.2C1-+ '-2- log [Cr]2 = EOFe++. Fe++++ -1- log [Fe++]'
N ow EOCI.,2CI- = 1·36 volts and EO Fe++. Fe+++ = 0·75 volts,
[Fe+;t+]2 X [Cl-]2 0·61
henc~ log [Fe++)2 X [CIJ = 0.02965 = 20'67 = l.og K,
or K = 4·7 X 1020.
The large value of the equilibrium constant- signifies that the
reaction will proceed from left to right almost to completion,
i.e., a ferrous salt is almost completely oxidised fly chlorine.
Consider now the more complex reaction:
Mn0 4- + 5Fe++ + 8H+ ~ Mn++ + 5Fei++ + 4H:p.
The equilibrium constant K is given by :
[Mn++] X [Fe+++] 5
K = [Mn0 4 ] X [Fe++]5 X [H+]8 '
The term 4H 2 0 is omitted since the reaction is carried out in
dilute solution, and the water concentration may be assumed
constant. The hydrogen ion concentration is taken,as molar.
The complete reaction may be divided into two half-cell
reactions corresponding to the partial equations:
Mn0 4- + 8H+ + 510 ~ Mn++ + 4H 2 0 (i),
and Fe++ ~ Fe'*"++ + 10 (ii).
For (i) as an oxidation-reduction electrode, we have:
E - EO . 0'059 1 [Mn0 4-] X [H+)B
- -t 5 og [Mn++] .,
_. 0'059 [Mn0 4-] X [H+]8
- 1 48 + 5 1og [Mn++] .
The partial equation (ii) may be multiplied by 5 in order to
balance (i) electrically:
5Fe++ ~ 5Fe+++ + 5 € '(ii'),
For (ii') as an oxidation-reduction electrode:
...
0·059 [Fe+++]5
E = EO + -5- 1og [Fe++] .5'
0·059 [Fe+++]5
0·75 + -5- log [Fe++]5 .
Combining the two electrodes in~ a ceJI, the e.m.f. will be
zero when equilibrium is attained: .
. 0'059 [Mn0 4 -] x [H+)B
'Le., 1·48 + -5- 1og [Mn++] .
0·059 [Fe+++] 5
:::'!:0'75 + -5- log [Fe++)S ,
log [Mn++] X [Fe+++)S _ 5(1·48-0·75) _ 61'!)
or [MnO] X [Fe++]5 X [H+]8 - . 0·059 - .
• 4
[Mn++] X [Fe+++]5 62
K = [MnO, ] X [Fe+++]5 X [H+]8 = 1·5 X 10 •
This result clearly indicates that the reaction proceeds vir-
tually to completion. It is a simple matter to calculate the
residual ferrous ion concentration in any particular case.
Thus suppose we titrate 10 m!. O'IN potassium permanganate
with approximately O'IN solution of ferrous ions in the pres-
ence of molar concentration of hydrogen ions. Let the volume
'of the solution at the equivalence point be 100 m!. Then
[Fe+++] == O'OIN, since it is known that the reaction is
practically complete, [Mn++] = 1/5 X [Fe+++] = O·002N.
and [Fe++] = x. Let the excess of permanganate solution
at the end point be one drop or 0'05 m!.; its concentration
will be 0·05 X 0·1/100 = 5·X 1O- 5N = [MnO,-]. Substitut-
ing these values in the equation:
(2 X 10- 3 ) X (1 X 10-.2)5
K = 5 5
- = 1·5 X 1062 •
(5 X 10- ) _X x X JB
or x = [Fe++] = 5 X lO-15N.
For the general oxidation-reduction reaction:
a OXr + b Red u ~ b OXu + a Red r (i).
Red a X OXII b
K= OXr'ar X Red b •
u
It can readily ,be shown that the concentrations at the stoi-
cl).iometric point when equivalent quantities of the two sub-
stances OXr a.nd 'Redn are allowed to react, are given by· :
Red l .. OXII = aHVR (ii).
OXI Red n
This expression enables us to calculate the exact concentration
at the equivalence point in any. redox reaction of the type (i),
and therefore the feasibility of a titration for quantitative
analysis. Consider again the interaction between, say, O'IN
solutions of potassium permanganate and a ferrous salt in the
presence of molar sulphuric acid. The equilibrium constant
of the reaction is 1·5 X ,1062 (p. 107). At the"equivalenc~
point :.
[Fe+++] HI .. 6 :::-:::-----:--=-=-::
= vl'5 X 1062 = \Ir'5 X 1062 = 2·3 X 1010 •
[Fe++] .
Since the ferric ion concentration will then be ca. o.,,05N, it is

evid~nt that for all practical purposes the reaction IS a quan-
titative one.
·It is clear from what has already been stafed that standard
oxidation potentials may be employed to determine whether
redox reactions are sufficiently complete for their possible use
in quantitative analysis. It must be emphasised, however,
that these calculations provide no information as to the speed
of the reaction, upon which the application of that reaction in
practice will ultimately depend. This question must form
the basis of a separate experimental study, which may include
the investigation of the influence of temperature, variations
of pH and of the concentrations of the reactants, and the
influence of catalysts. Thus, theoretically, potassium per-
manganate should quantitatively oxidise oxalic acjd in
aqueous solution. It is found, however, that the reaction is
extremely slow at the ordinary temperature, but is more rapid
at about 80°C, and also increases in velocity when a little
manganous ion has been formed, the latter apparently acting
as a catalyst.
It is of interest to consider the calculation of the equilibrium
consta?t of the general redox reaction (i), viz. :
a OXr + b Redn .= b OXn + a Re~I'
• See for example, Kolthoff and F\lnnan, Potentiometric Titrations, 1926,
p. 65 (John Wiley).
The complete reaction may be r~garded as composed of two
oxidation-reduction electrodes, aOxI, aRed I and bOxu,
bRed u , combined together into a cell; at equilibrium., the
potentials of both electrodes are the same:
E - EO 0·0591 1 [o'xI]a
1 - 1 + - n - og [.Redrl a '
E - EO 0'059J 1 [OXr]b,.
2 - 2 + n og [Redn]b'
At equilibrium, E 1 = E z, hence:
EO 0'0591 [OXI]a _ EO 0·0591 [Oxn]b.
1
1 + - n - og [Redr]a - '2 + --n- log [Redu]b '
[OXU]b X [Redrl a n
or log [Redu]b X [Oxr]a = log K = 0.0591 (E~ - E~) (iii).
This equation may be employed to compute the equilibrium

constant of any rep-ox reaction, provided the two standard
potentials E; and E; are known; from the value of K thus
obtained, the feasibility of the r~action in analysis may be
ascertained.
I, 49. Change of the electrode potential during the titration
of an oxidant and a reductant. Oxidation-reduction curve.-
In Sections I. 33 to 1,36 we calculated the change in pH during
the titration of an acid and a base, and we utilised the neu-
tralisation curve thus obtained in order (a) to specify the most
suitable indicator for use in the process, and (b) to compute
the titration error 'in the quant~tative neutralisation. We
shall now attempt to carry out parallel calculations for
oxidation-reduction processes. Let us study first a simple
case which involves only the valency change of ions, and is
theoretically independent of the hydrogen ion concentration.
A suitable example, for purposes of illustration, is the titration
of 100 ml. of 0·· IN ferrous iron with O·lN ceric cerium:
Ce++++ + Fe++ ~ Ce+++ + Fe+++.
The quantity corresponding to pH is the ratio [Ox] / [Red].
We are concerned here with two systems, the ferrous-ferric
ion electrode (i) and the cerous-ceric ion electrode (ii).
For (i) at-25°C.
0·0591 [Fe+++] [Fe+++]
El =E~+ - 1 - log [Fe++] = + 0·75+ 0·0591 log [Fe++] "
no Qu,antitative Inorganic Analysis
For (ii), at 25°C:
0·0591 [Ce++++]
E: = E~ -+ --I-log [Ce+++J
[Ce++++]
= + 1·45 + 0·0591 log [Ce+++] .
The equilibrium constant of the reaction is given by

(Section I, 48) :
[Ce+++] X [Fe+++]
log K - ,
- [Ce++++] X [Fe++]
I
= - - (1'45 - 0'75)
0·0591 .
= 11'84,
or K = 7 X 1011.
The reaction is therefore virtually comple,te.

During the addition of the cerie solution up to the equiva-
lence point, its only effect· will be to oxidise the ferrous iron
(since K is large) and consequently change the ratio [Fe+++]/
[Fe++]. When 10 mI. of the oXIdising agent have been added,
[Fe+++] I [Fe++] = 10/90 (approx.),
and E,=0·75+0·0591 log 10/90=0·75+0·056 '" 0·69 volt.
-With 50 m}. of the oxidising agent, E 1 = E~ = 0·75 volt.
With 90 mI., El = 0·75 -+ 0·0591 log 90/10 = 0·81 volt.
With 99 mI., El = 0·75 = 0'0591 log'99f1 = 0·87 volt.
With 99·9 mI., E} = 0·75 -+ 0·0591 log 99-9/0'1=0-93 volt.
At the equivalence point (100'0 m!.) [Fe+++] = [Ce+++] and
[Ce++++] = [Fe++], and the electrode potential is given by* :
E~ -+ E; = 0:75 + 1·45 = 1.10 volts.
2 2
,.. For a de~luction of this expression and a discussion of the approximations
involved, see Kolthoff and Furman, Potentiometric Titrations. 192~ p. 109.
It can similarly be shown (loe. cit. p. 107) that for the reaction:
a OXr + b RedIl ? b OXIl + a RedI.
the pJ;ltential at the equivalence point is given ?y :
bE~ aE~ +
E. = ---~-
a + b
where E~ refers to Ox,. Red,. and E~ to OXII. RedIl.
The Theoretical Basis of Quantitative Analysis III
The subsequent addition of the ceric solution will merely
increase the ratio [Ce++++] / [te+++]. Thus:
with 100·1 mI., El = 1·45 + 0'0591 log 0'1/100=1'27 volts;
with 101 mI., El 1·45 + 0'059~ log 1/100 = 1·33 volts;
>
with 110 mI., E 1 = 1·45 + 0·051)1 log 10/100 = 1·39 volts;

with 190mI., E 1 = 1·45 +0·05911og90/100:......1·45volts.
These results are plotted graphically in Fig. 1-16.
1·1S0
'EltJss Ce+->H
~
0
y
0
1'20
0 .P.
0
0·90
0·8 .... ) I.
0·70
'E ~
C? 20 40 60 80 100 120 140 180 180 200
0·7 N Cerie Sulphate added. mI.
FIG. 1-16. Titration of !OO ml. of O·IN ferrous iron with O'lN ceric sulphate
(calculated).
"It is of interest to calculate the ferrous iron concentration
in the neighbourhood of the equivalent point. When 99·9 ml.
of the ceric solution have been added, [Fe++] = 0·1 X 0·1/
199'9 = 5 X 10-5, or pFe++ = 4·3. The concentration at the·
equivalence point is given by (Section I, 48) :
. [Fe+++] / [Fe++] = K = 7 X 1011: = 8'5 X lOs.
Now [Fe+++] = 0·05N, hence [Fe++] = 5 X .10- 2/8'5 X 105 =
6 X lO-sN, or pFe++ = 7·2. Upon the addition'of 100·1 ml.
of eerie solution,- the oxidation potential (vide supra) is 1·27
volts. _The [Fe+++~ is practically unchanged at 5 X 10-2N,
and we may calculate [Fe++~ with sufficient accuracy for our
purpose from the equations :
[Fe+++]
E = E~ + 0'0591 log [Fe++] ,
5- X 10- 2
1·27 = 0·75 + 0·0591 log [Fe++]
[Fe++] = 1 = 10-1 °, or pFe++ = 10.
ThusPF e++ changes from 4' 3 to 10 between 0,1 per cen t before
to 0-1 per cent after the stoichiometric point. These quan-
tities are of importance in connexion with the use of indicators
for the detection of the equivalence point.
It is evident that the abrupt change of the potential in the
neighbourhood of the equivalence point is dependent upon the
,.7 standatd potentials of 'the
I ~ two oxidation - reduction
ce)H/ Ce",,"+-t+
systems that are involved,
1'5
Mn ·+/ Mn
0.-
~ I-- l
and therefore upon the
1'3
!/
l/ equilibrium
reaction; it
constant of the.
is independent
2Cr· ... + Cr,01-
1/
of the concentrations 1;lnless
,
.....
,. these are extremely small.
...: 1·0 The change in oxidation
.~
C o·g potential for a number of
~
2
cl: 0'8
'" 0·7
"... 0'6 1/
:l:!
.. -
Fe++j Fe++'"
~ typical oxidation-reduction
s y s t ems is e x h i bit e d
graphically in Fig. 1-17.
FO.r the MnOr, Mn++
<:> 0-5 system and others which
are dependent upon the
1)0'3 f- Methylene. hite/Methyl ne blue ~ PH of the solution, the
I hydrogen ion co:qcentration
0·2
'" is assumed tt> -be molar:
+0· 1
1"'f1'+++j Ti ++ + _v lower acidities give lower
0'0
,-,~F- potentials. The value at
1./
-0· 0 10 20 30 40 50 60 70 80 90100 50 per cent oxidiseCl form
Per cen't Oxidi8ed Farm will of course correspond
FIG. 1-17. to the standard oxidation
potential. As an indication
of the application of the curves, we may take the titration ·of
ferrous iron with potassium dichromate. The titration curve
would follow that of the ferrous-ferric system until the end point
was reached, then it would rise steeply and continue along
the curve for the dichromate-chromic system: the p9tential
at the equivalence point can be computed as already des-
cribed (p. llO).
I, 50. Indicators for the detection of the end point in
oxidation-reduction titrations.-A. Internal oxidation-reduction
indicators. We have already seen (Sections 1;32-34) that
acid-base indicators are employed to mark the sudden change
in PH during acid-base titrations. Similarly an oxidation-
The Theoretical Basis of Quantitative Analysis llS
reduction indicator should mark the sudden change in the
oxidation potential in the neighbourhood of the equivalence
point in an oxidation-reduction titration. The ideal oxidation-
reduction indicator will be one with an oxidation potential
intermediate between that -of. the solution titrated and that
of the titrant, and which ex~ibits a sharp, readily detectable
colour change. The subject is still in its infancy, and thus
far only a few such indicators have been discovered .
.An oxidation-reduction indicator is a compound which has
different colours in the oxidised and reduced forms:
lnox + nE ~ InRed'
The oxidation and reduction of the indicator should be
reversible. Many organic dyestuffs belong to this class; the
leuco compounds obtained on reduction are usually colourless,
and are converted by oxidation into the coloured dyes. If
such an indiGator is ?-dded to any redox system, for example,
one containing ferric and ferrous ions, the following equili-
brium will be present: .
Fe+++ + InRed ~ Fe++ + lilox .
If the quantity of indicator. added _is small, it will have little
effect on the main ferric-ferrous system, and it will adjust
itself so that its potential is the same as that of the solution
in which it is placed. The colour of the indicator system will
depend upon this potential. When the .concentrations. of
lnos and lnRed are equal, the colour will be intermediate
between those of the two extreme states, and the· potential
is equal to the standard potential of the indicator EO rn . It
can be shown that if the potential is about 0·05 volt greater
than EO rn' the indicatbr will be about 90 per cent oxidised and
its colour will be indistinguishable to the eye from that of the
oxidised form (compare Section I, 27); if the potential is
(EO rn - 0'05) volt, then about 90 per cent of the indicator
will be in the reduced form and it will exhibit its almost com-
pletely reduced colour. Thus the colour change for a redox
indicator will be over the pofential range (EO rn - 0·05)
to (EO rn - 0'05) volts. For a sharp colour change at the
end point, EO rn should differ by at least 0'15 volt from the
other systems involved in the reaction.
One of the best oxidation-reduction indicators is the ortho-
phenanthroline ferrous ion. The base brtho-phenanthroline
combines readily in solution with ferrous salts in the molecular
ratio 3 base: I ferrous ion 'forming the intensely red tri-
114 Qu:antitq,.t£ve Inorganic Analysis
ortho-phenanthroline ferrous ion; with strong oxidising
agents the ferric complex ion is formed, which has a pale blue
colour. The colour change is a very striking one.
[Fe(C 12 H sN2)a]++ ~ [Fe(C.12H8N~)3]+++ + E.
deep red pale blue
The oxidation potential is 1·14 volts. The indicator is of
great value in the titration inter alia of ferrous saIts and other
sqbstances with eerie sulphate solutions. It is prepared by
dissolving ortho-phenanthroline hydrate (molecular weight ='
198·1) in the calculated quantity of O'02M acid-free ferrous
sulphate, and is therefore tri-ortho-phenanthroline ferrous
sulphate. One drop is usually sufficient in a titration; this
is equivalent to less than 0'01 ml. of O·IN oxidising agent, and
hence the indicator blank is negligible. The only disadvan-
tage of this indicator is that it is somewhat expensive. It
may be replaced in many titratiolls by the le~s costly N-
phenylanthranilic aci.d (see Section m, 73), which has an oxida-
tion potential of 1·08 volts. Xylene cyanol FF (EOln > 1·0
volt) may alsa be e.mployed.
Mention should be made of one of the earliest internal
indicators. This is a one per ce~t solution of diphenylamine
in concentrated sulphuric acid, and was introduced by Knop
(1924) for the titration of ferrous salts with potassium di-
chromate solution. To obtain a satisfactory end point (an
intense blue-violet colollration), phosphoric acid must be
present; this forms a complex with the ferric ions, thereby
reducing the concentration of the latter, and consequently the
actual potential of the ferric-ferrous system is reduced well
below the normal value of 0·75 volt. The action of dipheny-
Q-NH-C) -t- Q-NH-C) ~ Q-NII-Q----O-.NH-G

(I) II) III)
-- Q-N=O===C)=N -0
1Il!)
lamine (I) as an indicator depends upon its oxidation first

into. colourless diphenylbenzidine (II), which is the real indi-
cator and is reversibly further oxidised to diphenylbenzidine-
violet (III). Under certain cOJ;}.ditions there is formed as an
intermediate product a very slightly soluble green compound
which is usually considered to be a molecular compound of
diphenylbenzidine and diphenylbenzidine-violet.
A solution- of diphenylbenzidine in conceptrated sulphuric
acid aets similarly to diphenylamine. The oxidation poten-
tial of the system II, III is 0·76 volt. It is therefore evident
that a lowering of the potential of the ferrous-ferric system is
necessary, as already mentioned, in order to obtain a sharp
colour change. The disadvantage of diphenylamine and of
diphenylbenzidine is their slight solubility in water. This
has been overcome by the use of the soluble barium or sodium
diphenylamine sulphonate, which are employed in 0·2 per
cent solution. The oxidation potential (EOln) is slightly
higher (D'83 volt), and the oxidised form.has a reddish-violet
colour resembling that of potassium permanganat~, but the
colour slowly disappears on standing; phosphoric acid must
st!ll be present in order to lower the oxidation potential of the
system. The addition of phosp'horic acid may be avoided by
the use of ortho-phenanthroline ferrous ion, N-phenylan-
thranilie aeid or xylene eyanol FF as indicators.
At this stage reference may be made to potential mediators,
i.e., substances whieh undergo reversible oxidation-reduction
and reach equilibrium rapidly. If we have a mixture of two
ions. say M++ and M+, which reaches equilibrium slowly with
an inert electrode, and a very small quantity of a eerie salt
is added, then the reaction:
M+ + Ce++++ -+ M++ + Ce+++
takes place until the tendency of M+ to be oxidised to M++ is
exactly balanced by the tendency of Ce+++ to be oxidised to
Ce++++, that is, until the M++, M+ and Ce++++, Ce+++ poten-
tials are equal. A platinum or other inert electrode rapidly
attains equilibrium with the cerous and eerie ions and will
soon register a stable potential which is also that due to the
M++ + € ~ M+ system. If the pot~ntial mediator is em-
ployed in small amount, then a negligible quantity of M+ is
converted into M++ when equilibrium is reached and the
measured potential may be regarded as that of the original
system. Potential mediators are of course useful in the
measurement of the oxidation-reduction potentials of redox
systems; in this eonnexion mention may be made of the use
of potassium iodide (= iodide-iodine system) in the arsenate-
arsenite system in acid solution and of eerie carbonate
(= ceric-cerous system). in the arsenate-arsenite system in
alkaline solution. It is evident that redox indicators (e.g.,
ortho-phenanthroline ferrous ion) may act as potential
mediators.
B. The reagent- may serve as its own indicator.. This is
well illustrated by potassium permanganate. One drop
(0,05 m!.) of excess of, say, O·lN potassium permanganate
will impart a visible pink-colouration to several hundred ml.
of solution, even in the presence of slightly coloured ions, such
as ferric ions. The colour of eerie sulphate and of iodine
solutions have also been ,employed in the detection of end
points, but ,the colour change is not so marked as for potas-
sium. permanganate; here, however, sensitive internal
indicators (ortho-phenanthroline ferrous ion or N-phenylar.-
thraniJic acid and starch respectively) are available. '
This metho~ has the drawback that an excess of oxidising
agent is always present at the end point. For work of the
highest accuracy, the indicator blank may be aetermined and
allowed for, or the error may be considerably reduced by
performing the standardisation and determination finder
similar experimental conditions.
C. External indicators. The best known example of an
external indicator in a redox process is the- spot test method
for the titration of ferrous iron with standard pota~sium di~hro
mate solution. N ear the equivalence point, drops'of the solution
are removed and brought into contact with dilute, freshly
prepared potassium ferricyanide solution on a spot plate.
The end point is reached when the first drop fails to give a
blue colouration. Another example is provided by the titra-
tion of zinc ions with standard potassium ferrocyanide solu-
tion; here a solution of uranyl acetate or nitrate is the exter-
nal indicator, and titration is continued until a drop of the
solution just imparts a brown colour to the indicator. Exter-
nal indicators' are gradually being superseded by the more
satisfactory internal oxidation-reduction indicators: thus in
the first example ortho-phenanthroline ferrous ion or N-
pheny!anthranilic acid .is suitable, whilst for the second
diphenylamine or diphenylbenzidine may be used.
·D. Potentiometric methods. These are discussed in the
following Section. They can usually be applied where no
satisfactory indicator is available.
I, 51. Potentiometric titration.-The various visual methods,
dependent largely upon the use o~ indicator!?, for the de-
tection of the end point in titrati9ns have already been
discusser.. We shall now deal wit:a a physico-chemical
method which may be applied not only to those cases where
suitable indicators are available, but also to those in which
the visual indicator methoQ. fails or is of limited accuracy,
e.g., for coloured or for very dil_ute solu~ions; the procedure
is- termed potentiometric titration. This method may .often
be employe-d for the determination of two or more constitu-
ents in a single operation. Thus a mixture of iodide and
chloride may be titrated with silver nitrate solution; the
first point of infie~ion is the equivalence point of the iodide
reaction and the second inflexion point is that for1he chloride
reaction (compare Fig. 1-12). .
The theory of potentiometric titration is very similar to
that already given for ordinary titrations. It has been
shown (Section 1.43) that the potential of a metal electrode and
a solution of its own ions at 25°C is given by the expression:
.E26 Q
= E ° + 0·0591
-n-Iog c.+ (i),
where EO is the standard potential of the metal, n fhe valency

of the ions, and Cj+ is the ionic concentration (strictly, this
should be the ionic activity). If we write Ci+ in the exponen-
tial form, i.e.,
- log [Me"+] = pMe"+, the equation becomes:
E 25 = EO _ 0'0591 pMe"+
0 (i'):'
n
The expression for a hydrogen electrode at 25°C is:
E 25 = EOH - 0·059J
0 pH (ii) ,
where EOH is the standard potential of the normal hydrogen
electrode. For an oxidation-reduction electrode, the expres-
sion is:
E 25 - EO + 0'0591
Q - [Ox]
n 1og [Red] (iii) ,
where EO is the standard oxidising potential, n the number of

electrons gained by the oxidant in being converted into the
reductant, and [Ox] and [Red] are the concen.trations (strictly,
the activities) of the oxidant and reductant respectively.
The determination of E is the ess~nce of potentiometric
titration: this may be used either to measure the ionic
(or [Ox] / [Red]) concentration of a solution, or, what is more
important for our present purpose, to follow the changes in
ionic concentration 01' of the ratio [Ox] / [Red1 during the
course of a titration.
Since EOH , the normal potential for hydrogen, viz., the
potential of the hydrogen electrode immersed in a solution of
hydrogen i.ons of unit activity at one atmosphere pressure,
is taken arbitrarily as zero at all temperatures, expressiQn
(ii) reduces to :
E = - 0·0591 pH (at 25°C*)
on this scale. On theoretical grounds, therefore, the deter-
mination of the hydrogen ion concentration is the simplest
case to deal with, and it is proposed to discuss this first.
I, 52. Determination of pH.~A. The hydrogen electrode. A
typical hydrogen electrode has already been described (Sec-
tion I, 43). For routine work, the most convenient source
of hy~rogen is the compressed gas, sold in cylinders; by
means of a suitable reducing valve, a steady, stream of hydro-
gen gas can be readily obtained. The gas should be passed
through all-glass wash
bottles containing respec- To Potentiometer
tively alkaline potassium t
permanganate sol uti 0 n,
a1kaline pyrogallol solution
and purified glass wool before
:reaching the electrqde. If
a hydrogen electrode is
immersed in a solution the
pH of which is to be
determined, and the' half
cell coupled with a molar ~[H+~l~ ':[I1":f~ ~-
hydrogen electrode by means FIG. 1-18.
of a potassium chloride
bridge (Fig. l-lSj· in order to eliminate the liquid junction
potential,t the e.m.f. of the resultant cell :
Pt I Hz, H+' II H 2 , H+ Pt,
a=l a=x
• At any other temperature T, this becomes:
E = - O·0001982T. pH,
where T is the absolute temperature.
t The open ends of the bridge are plugged tightly with pure filter paper in
order to minimise diffusion effects. Alternatively, a jelly of 3 per cent agar
in saturated potassium chloride solution-may be used.
may be measured by means of a potentiometer (see below).
The e.m.f. of this concentration cell (see Section 1,44) at 25°C
is given by : ..
-_ 0·0591 1 c2
E - - - og-,
n c1
0·0591 1
--1-log [H+] ,
- 0·0591 pH;
or pH = E/0·0591.
The use of the molar hydrogen electrode as a standard
electrode or half cell presents certa~~ practical difficulties.
It is usual to employ some form of the calomel electrode as a
secondary standard half-cell. A' calomel electrode is one in
which mercury and calomel are covered with potassium
chloride of definite concentration; this. may be O'IN, N,
3'5N or saturated. The potassium chlorid~ solution must be
saturated with calomet-. The potential of, say, the saturated
calomel electrode (i.e., that prepared with saturated potassium
chloride solution) must' first be determined with reference to
the molar hydrogen electrode. Let this value be designated
Ecal.•m.' For the measurement of hydrogen ion co~centra-
tion, we then employ the cell: • .
Hg I Hgll 2 , KCI (satd.) II H+, 'H! I Pt.
Let the resultant e.m.f. be E 06s .• Then, at 25°C.:
E obs• = E cal• sat.·~ 0·0591 log [H+] ;
pH = E obs• - E cal• s• .
0·0591 '.
The valUe of E cal . sm • is 0·246 volt at 25°C.

Various forms of the calomel electrode are illustrated in
Fig: 1-19. One of these, (a), will be described in detail; the
others will then be self-evident. It consists of a glass vessel
provided with a bent side tUDe A and another side tube B.
over the end of which a piece of rubber tubing is placed which
can be closed by a spring or screw clip. Electrical connection
with the electrode is made by means of a platinum wire.
sealed through a glass tube C; the latter contains a little
pure mercury into which an amalgamated copper wire dips.
To set up the electrode. a saturated solution of analytically
pure potassium chloride containing some of the solid salt is
PotallSJum
chlorld.
solution
Calomel
paste
Mercur!/
(a) (b) ra)
FIG. 1-19.
first prepared. Pure mercury to a depth of 0·5 to 1 cm. is

placed in the bottom of the dry electrode vessel; the mercury
is then covered with a layer of calomel paste D. -The latter
is prepared by rubbing pure calomel, :p1ercury and saturated
potassium chlorid~ solution together in a glass mortar;
tIle supernatant liquid, is poured off and the rubbing process
repeated twice with fresh.quantities ,of saturated potassium
chloride solution. The rubber bung carrying the plass tube
and platinum wire is then inserted, care being taken that the
platinum wire dips into the mercury. The vessel is then filled
with a saturated solution of potassium chloride (previously
saturated with calomel by shaking with the solid salt) by
drav'ing in the solution through the bent tube A, and then
closing the rubber tube B with a- clip. The electrode is then
ready for use. In electrode (b), the siphon tube or salt
bridge may be filled with a jelly of 3 per cent agar in saturated
potassium chloride solution. The electrode (c) is suitable
for precision work; it has a 3-way stop-cock for flushing
away the contaminated potassium chloride after it has been
employed in a titration. .
The most satisfactory method for the measurement of the
e.m.f. of a cell is that known as Poggendorff's compensation
method, an outline of which is given below." The principle
of the method is to balance the unknown e.m.f. against p.
known e.m.f., which can easily be varied. When these two
e.m.f.s are exactly equal, no current will flow tb.rough a
galvanometer pl~ced in the circuit: the.. galvanometer is
therefore employed as a null instrument. The essential
details are shown in Fig. 1-20.
The' Theoretical Basis oj Quantitative Analysis 121
A 2- or 4-volt Rheostat
accumulator furnishes
the opposing e.m.f.;
this is connected in c
series with a rheostat AI-------.....__-..:r---_}B
and with the terminals
of a slide wire AB. The
latter is a thin wire of
uniform cross section
and is often termed 'the
"potentiometer wire."
The cell, the e.m.f. of
which is to be deter- Cell
mined, is connected to FIG; 1-20.
one end A of the slide
wire, and through a galvanometer G 'and a' key 8 2 to a
sliding contact C, which can be moved along AB, A
special double-throw switch 8 1 may be provided to permit
the standard cell to be placed in the circuit. In connecting
the accumulator and the cell to the bridge, it is essential
that the positive poles should be connected to the same
end of the bridge wire; the unknown cell will then send
a current through the circuit in a direction opposite to that
furnished by the accumulator. •
If we assume that the potentiometer wire has uniform cross
section and resistance, then the fall of potential along the
. slide wire will pe uniform. The difference of potential
between A and any point C will be p:lmportional to the length
AC, and will be equal to the fraction ACjAB of the total fall
of potential along the wire. If the unknown cell is'now placed
in circuit and the position of C so adjusted that when the
switch 82 is depressed, no current passes through the galvano-
meter G. then the e.m.f. of the cell is equal to that of the
accumulator multiplied.by ACjAB. For most potentiometric
work, only changes of potential are required, so that variations
of the length AC are all that are required during a titration.
In general, however, the e.m.f. of the accumulator is. not
guite constant. A standard cell is employed. This is usually
a Weston cell. which has an e.m.f. of 1,0183 volts at 20° e or
1·0183 - 0·0000406 {t - 20°)
at any other temperature tOe. If the standard cell is placed

in circuit by means ~f the switch 8 1 and the point of balance
C' on the bri~ge is determined, then the unkno~ e.m.f.
may be calculated from the relation':
AC e.m.f. of unknown cell
AC' e.m.f. of standard cell
For approximate work, the slide wire AB may consist of an
ordinary meter bridge, and the indicating instrument may be
a milli-ammeter. It is preferable, however, to employ a
commercial type of potentiometer, which utilise the more
compact spiral type of bridge wire. * The most convenient
type of indicating instrument is the direct vision type of
mirror galvanometer; the galvanometer and lamp scale are
incorporated in a blackened wooden box and the "spot"
is clearly visible in daylight,t The rheostat in the commercial
potentiometer provides for the drop of potential along the
slide wire. If the bridge is divided into, say, 2000 equal parts,
then the rheostat may be adjusted with the standard cell in
circuit and with the sliding contact C at a position corres-
ponding to 1018·3 divisions so that no current flows through
the galvanometer. The position of the sliding contact will
then give the e.m.f. of any unknown cell -directly in milli-
volts. It is usual for the rheostat to contain both coarse
and fine adjustments: the fine adjustment may. be used to
compensate for the slight variations of the accumulator during
the measurements.
The hydrogen electrode cannot be used in solutions contain-
ing oxidising agents, e.g., permanganate, nitrate, ceric and
ferric ions, or of other substances capable of reduction such
as unsaturated organic compounds, or in the presence of
sulphides, compounds of arsenic etc. (catalytic poisons)
which destroy the catalytic property of platinum black. It is
also unsatisfactory in the presence of salts of the noble metals"
e.g., copper, silver and gold, and also in solutions containing
lead, cadmium and thallous salts. There are many other
electrodes which~ are more convenient to use in the range in
which they are applicable. Some of these will' be described
below.
• Excellent potentiometers are made by H. Tinsley and Co. of London, The
Cambridge Instrument Co. of London, and the Leeds and Northrup Co. of
Philadelphia, U.S.A. Upon r&quest, these firms will supply pamphlets con-
taining full details, including the wiring circuit. of their instruments.
t The, author has found that marketed by H. Tinsley and Co. of London
very. convenient in practice; many other similar types'are available commer-
cially, .
The. Theoretical Basis of Quantitative Analysis 123
B. The quinhydrone electrode:' The preparation of repro-
ducible hydrogen electrodes is a comparatively tedious pro-
cess and, moreover, the electrode has certain limitations (see
previous paragraph). E. Biilmann (1921) has introduced the
quinhydrone electrode: this..,renders the determination of pH
a rapid and simple process, and does not require the use of
hydrogen gas. Quinhydrone is a compound of quinone and
hydroquinone. and in solution is decomposed into equimole-
cular quantities of these substances:
C6H402,C6H4(OH)z = C6 H 4 0 2 + C6 H 4 (OH)z'
quinhydrone quinone hydroquinone
Quinone and hyciroquinone form a reversible oxidation-
reduction system which may be represented as:
C 6H,02 + 2H+ + 2.: ~ C6 H 4 (OHh.
If an inert .electrode', such as platinum, is immersed in the
system, the potential is given by (Section I, 47) :
RT a(2 • aa+ 2
E = Eo + 2F log. '
aHaQ
RT I aHzQ RT
= Eo. + 2F oge -a- + F logeaH+,
Q
where aQ' aH+ and aHzQ are the activities of the quinone,
hydrogen ions and hydroquinone respectively, and Eo is the
standard potential referred to the molar hydrogen electrode.
Now under those conditions in which quinhydrone dissociates
to give equimolecular quantities of quinone and hydroquinone,
the ratio of tne activities may be regarded as constant, hence:
RT
E = Eo + -p loge aH+ (i).
Eo has been determined in the usual manner by direct refer-
ence to the molar hydrogen electrode, a.nd has a value of
0·7044 volt at 18°C and 0·6991 volt at 25°C. By making the
uSllal substitution for the known values of the constants in
equation (i), we have, at 25°C: '
E = 0·69"91 + 0'0591 log aH't (i.:).
Thus the potential of the quinhydrone electrode changes with
the hydrogen ion activity in a manner which is exactly similar
to that of the hydrogen electrode.
For convenience, a calomel electrode ofren replaces the
molar hydrogen electl"ode as the other half cell. The complete
cell is then :
Hg f HgzClz, KCI (satd.) II Solution, quinhydrone I Pt.
The potential of the saturated calomel electrode, E eal . sat.'
against the molar hydrogen electrode is 0·2458 volt at 25° C.
Hence E obs . ~ Equinhydron. - E eal. sat.,
= 0·6991 + 0'0591 log [H+] + 0'2458,
= 0·4533 + 0'0591 log [H+].
PH = -log [H+] = 0.453:'~9~ObS'. (ii) .
To carry out ,a determination of the hyijrogen ion concen-

tration of a solution, about one gram of quinhydrone* per 100
ml. of solution is added, and the solution stirred. A bright
platinum electrode is immersed in the solution and the quin~
,hydrone electrode combined in a cell witl;l a saturated calomel
electrode; a saturated solution of potassium chloride or, less
accurately but perhaps more conveniently, a jelly of 3 per~ent
agar in saturated potassium chloride solution may be used as
tlte salt bridge. The PH is then calculated by equation (ti).
It may be mentioned thfl,t the platinum electrode should be
highly polished and cl~an; it is best cleaned.by treatment
with hot chromic acid mixture, then thoroughly washed with
distilled water, and ignited in an alcohol flame.
An alternative method is to use a solution of known pH as
standard and to combine the two quinhydrone electrodes into
a concentration cell :
Pt \ pSoHlutio~ hof udnknown II SpOlHutioh.ofhkdnown Pt.
,qUIll Y rone ,qUIll Y rQn"e
Then E obs . = 0.0:91 log ~ at 25 C,
0
CI
where C2 and cl,are the respective hydrogen ion concentra-
tions;
,
where pH 2 refers to C2 and pHI to ((1' If PH 2 is known, pHI
can be readily calculated.
The advantages of the quinhydrone electrode are: (i) it
attains equilibrium rapidly; (ii) it may be used in many
solutions in which the hydrogen electrode is inapplicable, e.g.,
in solutions of 'zinc, cadmium, tin, lead, ~opper and nickel
.* Pure quinhydrone .(analytical reagent quality) is available commercially.
It can also be readily prepared by oxidation of hydroquinone with ferric
ammonium sulphate at 65° C.
salts, dilute nitric acid, and in solutions of unsaturated organic
acids, of amino acids and of alkaloids; (iii) it is not so readily
incapacitated as the hydrogen electroqe. Its disadvantagel?
are: (i) it cannot be used in solutions of pH greater than
about pH 8 (this is because hydroquinone behaves as a weak
dibasic acid and then begins_to have an effect on the pH of the
solution; moreover, this will also affect the molecular ratio
of quinone to hydroquinone since the concentration of the
undissociated hydroquinone becomes reduced); (ii) the poten-
tial may be slightly affected by large concentrations of
neutral salts-the "salt error": this error, however, is
negligible under the conditions that the electrode is normally
used. .
.C. The glass electrode. F. Haber and Z. Kleinensiewicz (1909)
found that the potential difference between a thin glass surface
and the solution in which it was immersed varied in a regular
manner with the pH of· the solution. This remarkable fact
has been utilised in recent years as the basis of a valuable
method for measuring the pH of solutions. Ordinary soft
glass has too great a resistance to be of any great value .• A
suitable glass has the composition 6 per
cent CaO, 22 per cent Na 20 and 72 per
cent Si0 2 ; this is the lowest melting
point glass which can be prepared
using these three ingredients. The A A
glass made by the Corning Company
of ~erica,. known as Corning No. 015
glass, is also suitable. Two types of
electrode are employed. In the first B
type (a in Fig. 1-21), a tube A of the (b)
special glass is blown out at the lower
B
end into an extremely thin bulb B. FIG. 1-21.
In the second type (b in Fig. 1-21),
due to MacInnes and Dole, a ,:"ery thin membrane of the
special glass is sealed on to the end of a glass tube,
which is smooth and square. This is accomplished by first
blowing a large thin bubble on the end of the special glass
tubing; the end of ~ is heated to redness and then pres~ed
against the thin b~lb. The heat in the tube A is sufficient
to fuse the thin membrane into place. Considerable experi-
ence is required to prepare satisfactory electrodes of type
(b); in general, particularly for routine work, the bulb type
is to be preferred. These are readily prepared and can also
be purchased. The glass electrode is filled to a depth of 1
to 2 cms. with O.IN· hydrochloric acid and a little quin-
hydrone or with a buffer mixture (such as sodium acetate-
acetic acid mixture), and a clean platinum wire sealed through
a glass tube is inserted to make electrical connexion. 'The
bulb is then immersed in a solution of which the pH is to be
measured; a calomel (Qr other.. convenient electrode) com-
pletes the cell. Owing to the high resistance (5-50 meg-
ohms) of the glass electrode, special methods must be adopted
for measuring the e:m.f. of the cell. Perhaps the most
convenient in practice is to employ an ordinary commercial
potentiometer coupled with an appropriate amplifying unit*
and a §uitable ballistic galvanometer. * Special precautions
must be taken to avoid electrical leakage, for example, the
upper part of the tube A should be waxed and, if possible, the
cell should be immersed in a thermostat containing a special
paraffin oil. When not in use, the electrodes. must pe kept
immersed in water. '
If the glass electrode described above is used, the potential
of the electrode at 25°C is given by :
E • K - 0·0591 pH,
where K is a constant; this expression holds up to a pH of
about 10. The value of K is most simply determined by
immersing the glass electrode in buffer solutions, coupling
with a calomel or hydrogen electrode and measuring the
e.m.f.; this standardisation must be carried out at least
once a day since K varies slightly. Variation from "this
expression occurs at higher PH values, but by calibration
with suitable buffer solutions, the glass electrode may be
employed up to a.pH of 12..
The advantages of the ~lass electrode are: (i) no chemical
is added to the solution; (ii) the e.m.f. is independent of
oxidation-reduction po~entials; (iii) it is not affected by the
colour or turbidity of solutions; (iv) it can be used in un-
buffered solutions; (v) its most useful range is pH 0 to )0,
which may be extended by careful calibration to pH 12;
(vi) it attains equilibrium comparatively rapidly.. Its dis-
• Such an amplifying unit is obtainable from H.Jinsley and Co., of Wemdee
Hall, London, England, as is also the ballistic galvanometer. A compre-
hensive pamphlet, embodying the circuit, etc., is supplied by this firm. In
recent years more compact, but necessarily less accurate, instruments have
been introduced (see Section I, 53D) ; these embody a thermionic valve poten-
tiometer, are generally of the portable type, and are claimed to give' an
accuracy of about 0·02 pH units. Very rugged glass, calomel, and other elec-
trodes have been developed for use with these portable instruments; these
ar. shown in photograph at bottom of Plate I.
advanta£es are: (i) the high electrical resistance of the glass
which necessitates special measuring equipment and also
special precautions to prevent electrical leakage; (ii). ·its
fragility (see, however, footnote above and also Plate I).
I, 53. Experimental details concerning potentiometric
titrations.-A. Neutralisation reactions. The simplest appara- '
tus consists of a beaker containing the base and into which is
immersed a hydroge~ or glass electrode, the apparatus being
connected as shown in Fig. 1-20. A burette containing the
acid is clamped immediately above the beaker, and s'ome form
of mechanical stirring is provided for the solution. If the
quinhydrone electrode is to be used, the acid is placed in the
beaker and the, base in the burette. The e.m.f. of the cell
containing the initial solution is determined, and then the
measurement is repeated at intervals of 2 mI. addition from
the burette. When the e.m.f. changes somewhat more
rapidly as the end point is approached, the e.m.f. readings
are taken at more 'frequent intervals, and eventually after
each drop. After the first rough titration, the exact proce-
dure will be apparent. -Draw a curve with the e.m.f. readings
(or PH) as ordinates and the burette readings as abscissae:
the end point of the break in the curve determines the
equivalence point. A more satisfactory determination of the
end point. is obtained by the differential method, in which the
change of e.m.f. (or PH)
per unit of titrant is the
ordinate. If we plot the
change of potential in
milli-volts caused by 0·1 t
m!. increments of the ::...
reagent (.dEIAV) against <l
the volume of the titrant ""'L------
correspondlng t o t h e <l
middle of the titration
interval (here. the burette ~ Volume. mi.
reading + 0'05 mI.), a FIG. 1-22.
curve of the type shown
in Fig. 1-22 is obtained. The abscissa corresponding to
the maximum gives the end point of the reactton with
sufficient accuracy for all ordinary purposes. It is possible
by means of the differential method of potentiometric titration,
introduced by D. C._Cox in 1925 and subsequently improved
by D. A. MacInnes, to obtain values of .dEI.tJV directly and
thus to considerably increase the accuracy of the~determina
tion of the end point. *
the accuracy with which the end point can be found poten-
tiometrically depends upon the magnitude of the change in
e.m.f. in the neighbourhood of the equivalence point, and this
depends upon the concentration and the strength of the acid
and alkali (compare Sections I, .33-36). Satisfactory results
are obtained ih all cases except (a) those in which either the
acid or the_base is very weak (K < 10- 8) and the so!utions are
dilute, and (b) those in which both the acid and the base are
weak. In the latter case an accuracy of about 1 per cent
may be obtained in O·lN solution. \
The method may be used to titrate a mixture of acids
which differ greatly in their strengths, e.g., acetic and hydro-
chlodc acids; the first break in the titration curve occurs
when the stronger of the two acids is neutralised, and the
second when neutralisation is complete. For this method
to be successful, the two acids or bases should differ in strength
by at least 105 to 1.
B. Oxidation-reduction reactions.-The apparatus employed

here is similar to that for neutralisation reactions. The
beaker contains the reducing agent, appro:griately acidi-"
fied, if necessary, and also a calomel·electrodl! and a bright
platinum electrode, similar to that used with the quinhydrone
electrode (Section I, 52B). The burette will contain the
oxidising agent. The experimental det:;til:s, including the
determination of 'the end point, are identical with those
already given.
Numerous oxidation-reduction titrations can be carried'
out in this manner. A few typical ones are enumerated
below·t
Potassium permanganate (acid) with ferrous i~on, oxalate
(at 70°C), ferrocyanide, iodide, nitrite, stannous tin, thallous
tha~lium, and titanous titanium.
Potassium dichromate (acid) with ferrous iron, ferrocyanide,

iodide (in Hel), antimonious antimony, arsenious arsenic,
stannou~ tin, and titanous titanium.
~. See, for example, D. A. MacInnes, The Principles of Electrochemislrj,

p. 306 (Rheinold'Publishing Corporation, 1939) ; W. W. Scott-No H. Furman.
Standard Methods of Chemical Analysis, Vol. II, p. 2318 (1939).
f For further details, see Kolthoff and Furman; Potentiometric TUrations,
1931 (J. Wiley and Co.).
The Theoretical Basis of Qttantitative Analysis 129
Gerie sulphate with ferrous iron, oxalate (catalyst present),
ferrocyanide, arsenious arsenic, antimonio1l.s antimony,
stannous tin, hydro_gen peroxide, nitrite, mercurous mercury,
thallous -thallium, titanous titanium, tellurite, vanadyl salts
and uranyl salts.
Iodine with thiosulphate, copper salts in the presence of
acetic acid, and arsenites in the presence of bicarbonate.
Titanous c'k~oride or'sulphate with ferric iron, antimonate,
f~rricyanide, m~rcuric mercury, stannic tin, titanic titanium,
and uranyl salts (55-60°C).
C. Precipitation. reactions. Here one electrode may be a
saturated calomel or, less conveniently, a hydrogen electrode,
and the other must consist of one which will readily come into
equilibrium with one of the ions of the precipitate. For
example, in the titration of silver. ions with a halogen (chloride,
bromide or iodide) this must be a silver electrode. It may
consist of a silver wire, or of a platinum wire or gauze plated
with silver. and sealed into a .glass tube. Since a halide is
to be determined, the salt bridge must be a saturated solution
of potassium nitrate. Excellent 'results are obtained by
f~trating silver nitrate solution with thiocyanate ions.
RECENT DEVELOPMENTS IN THE TECHNIQUE OF POTEN1I-

OMETRIC TITRATION AND OF OTHER PHYSICAL METHODS
OF ANALYSIS. .
D. Potentiometric Titration. The use of the troublesome refer-
ence electrode and salt bridge was first eliminated through the
development of the polarised mono-metallic system and:of the
bi-metallic self-polarising system. The former consists of two
pure platinum wires, one polarised anodically and the other
cathodically with a polarising current of 0·5 X 10-5 ampere
(Willard and Fenwick, 1922). Such a system gives a sharp break
of potential at the end point of many, particularly oxidation-
reduction, titrations. This is due to the difference in
... This is ,most conveniently carried out by immersing the platinum in O'lN
silver nitrate to which enough potassium cyanide has been added just to
dissolve the I?recipitate initially fomed. Electrolyse the solution of potassium
argentocyamde, using the platinum to be plated as the cathode and a platinum
wire as anode. A 6-volt accumulator in series with a variable resistance may
be used, and the current adjusted to about 0·05 ampere. The plating i.!l
continued for about 15 to 20 minutes when a fairly thick deposit of silver
should be formed. The electrode must be thoroughly washed to remove all
traces of cyanide.
F
'the degree of reversibility of the electrode reaction before
and after the end point. The action of bi-metallic
systems will be readily understood from the following
discussion.
It has been found that certain metals and non-metals,
e.g., natural graphite, carborundum, silicon, cadmium,
nickel, cobalt, and copper, acquire a more or less definite
potential when immersed in solutions,. and that these do not
~hange appreciably as the pH' of the solution in which they
are placed is altered. On the other hand elements, such as
platinum, antimony and tungsten, adjust themselves more or
less in a regular manner as the PH of the solution in which
,they are immersed is varied. If one of the former type of
electrode is combined with one of the latter, then there is a
small difference of potential between them when they are
placed in, say, an acid solution which is to be titrated.
During the progress of the titration there is little change in
the e.m.f. of the·cell, but at the end, point the potential of one
electrode challges rapidly whereas that of the oth~r remains.
approximately constant. Thus there' is a sudden change of
e.m.f. at the equivalence point. For oxidation-reduction
titrations it has been shown that palladium and other metals
of its family in the periodic system (but excluding platinum)
and also tungsten do not change appreciably in potential as
the concentrations of the oxidant and reductant are varied.
The most useful electrode pair for this purpose is platinum-
• tungsten. The potential of the former responds readily to
changes in concentration of the oxidant and reductant, but
that of the-latter does not: at the end of the titration there
is a large increase in the e.m.f. of the Pt-W cell.
Subsequent improvements have led to the elimination of the
potentiometer, of the tapping key, and of the storage battery.
This was rendered possible by the use of thermionic valves;
the current drawn from the titration ceU can be amplified
without polarisation. The" thermionic titrimeter without
batteries" of Willard and Hager* (1936) is of this type; it
employs a Weston micro-ammeter as the indicating instru-
ment. G. F. Smith (1936) has designed a similar apparatus
which, however, replaces the sluggish, fragile and expensive
micro-ammeter by a small cathode-ray tube (" magic eye "),
which is inexpensive and highly sensitive. He terms the
* This apparatus is marketed by E. H. Sargent and Company. 155 East
Superior Street. Chicago. U.S.A.
PLATE r
FIG . 1- 33
FIG. 1-'24
CELL COMPLETE .
apparatus a "sectromet9r"*; this will doubtless find very
wide application when its properties and advantages are fully
Tealised. The apparatus is placed along the side of the
titration vessel, and its effect is somewhat comparable to the
use of a colour indicator; the only differences are a greater
warning of the approaching end point and the instantaneous
response of the cathode ray tube. A wink of the" magic
eye" indicates the completeness of the reaction.
It must, however, be pointed out that the first real advance
in potentiometric technique was made by K. H. Goode in
1922, who replaced the usual potentiometer by a thermionic
valve and thus introduced what are now commonly termed
thermionic valve potentiometers.t Numerous improvements
have subsequently been made, chiefly with the object of
utilising the glass electrode. An excellent commercial
instrument is the Cambridge pH meteq; the instrument
itself is shown in Fig. 1-23, and the circuit and connexion in
Fig. 1-24 (see Plate I). The instrument may be employed with
the glass, hydrogen, quinhydrone, and antimony electrodes,
• The complete .. sectrometer," including electrodes, is supplied by the
G. F. Smith Chemical Co., of 867 McKinley Avenue, Columbus, Ohio, U.S.A.
The agents for Great Britain are F. W. Berk and Co. Ltd., 52-54 Leadenhall
Street, London, E.C.3, England. The apparatus is designed for use on 110
volts and 60 cycles A.C., but by connecting a 25 watt 110 volt lamp in series
with the sectrometer circuit it may be employed with 220/240 volts 50 cycles
mains. A very comprehensive booklet entitled" The Electron Beam Sectro-
meter for Potentiometric Tritrations .. is available.
The Fisher Scientific Company of 711-723 Forbes Street, Pittsburgh. Pa .•
U.S.A .• market what.. they call the" Fisher titrimeter." This somewhat
elaborate apparatus also incorporates the" magic eye" as indicator, and may
be used for acid-base, oxidation-reduction, and conductivity titrations. and
also for measurements of D.C. voltage and resistance. It is, of necessity. more
expensive than the Smith" sectrometer ".
t For further details, see, for example. S. Glasstone, Electrometric Methods
in Physical and A nalytual Chemistry (Institute of Chemistry, London. 1934) ;
H. T. S. Britton, Hydrogen Ions, 2nd Edn., p. 133 (Chapman and Hall, 1932) ;
N. A. McKenna, Theoretical Electrochemistry, p. 41 (Macmillan, 1939); W. W.
Scott-No H. FUrman, Standard Methods of Chemical Analysis, Vol. II, p. 2321
(Van Nostrand-The Technical Press, 1939).
! Figs. 1-23 and 1-24 are reproduced by courtesy of the Cambridge Instru-
ment Co. Ltd., of 13 Grosvenor Place, London, S,W.I, England. This firm
also supplies a Can;tbridge Portable pH Meter, which is self-contained but in
other respects is similar to the above. Numerous other pH meters are mar-
keted; tpese include; the Portable Universal pH Indicator (Leeds and
Northrup Co., Philadelphia. Pa., U.S.A . and Integra Co. Ltd., Birmingham),
the Beckman pH Meter (Eimer and Amend, New York, U.S.A., and Griffin
and Tatlock Ltd., London) ; the Coleman pH Electrometer (Coleman Electric
Co., Maywood. Ill.. U.S.A .. ~nd Baird and Tatlock (London) Ltd ., Landon) ;
Hellige pH Meter (Hellige Inc., Long Island City, N.Y., U.S.A., and A.
Gallenkamp and Co. Ltd., London). Suitable glass and calomel electrodes are
supplied inter alia by the Cambridge Instrument Co., and are shown at bottom
of Plate I.
and readings may be obta:ined ei,ther directly in pH units (with
the glass electrode and towithi~ 0'01-0·02 pH units) or in milli-
volts (to within 0·001-0·002 volts) ; it can also be adapted for
redox and precipitation titrations. For fUTther particulars,
the reader is referred to the painphle.ts supplied by the manu-
facturers.
E. Conductometric *itrations. A diss:;ussion of this method,
which depends upon the measurement of the electrical con-
ductivity of the solution after successive additions -of the
reagent, is outside the scope of this book. The. reader is
referred to ~he various special monographs devoted to the
subject.* Mention must, however, be made of the fact that
the cathode ray tube (" magic eye ") has been incorporated in
a r.ommercial apparatust for the rapid and comparatively
a.:curate measurement of the conductivity of solutions; this
is particularly useful for conductometric titrations.
F. Spectroscopic methods. Methods of quantitative analy-
sis dependent upon the use of spectroscopy are finding increas-
ing application, particularly in metallurgical analysis. Their
great disadvantage is that an expensive equipment is required,
but when once this has been obtained results can, in many
cases, be obtained with comparative rapidity.=!:
G. Polarographic methods. By applying a steadily increas-
ing e.m.f. to an electrolytic cell with a dropping mercury
cathode and a stationary mercury p,node (the latter is covered
by the solution to be analysed) and utilising a device whereby
the current and voltage are measured and automatically
recorded on a revolving drum, current voltage curves may be
obtained. Such an :instrument is termed a polarograph and
the corresponding current-voltage, diagram is called a polaro-
gram. When the solution contains a substance which can
be reduced, e.g., a metallic ion or any other oxidising agent
(inorganic or organic), a sudden increase in current is observed
at the voltage at which reduction occurs (decomposition
... See, for example, H. T. S. Britton, Conductometric Titralitms, 1934
(Chapman and Hall). •
t This is termed the" Mullard Measuring Bridge, Type GM, 4140" and is
manufactured by the Mullard Wireless Service Co. Ltd., of 225 Tottenham
Court Road, London, W.1, England.
t For further details see, for example, Thorpe's Dictionary of Applied Chem-
istry, Volume II, p. 688 (Longmans, Green and Co., 1938); A. C. Candler,
SPeclograPhic Analysis in Great Britain (Adam Hilger, 1939); W. R. Brode,
Spectrographic Analysis in Scott-Furman, Standard Methods of Chemi<;al
Analysis, Vol, II. p. 2592 (1939).
The Theoretical Basis of Qttantitative Analysis 133
;voltage, see Section I, 65) ;"" this voltage is more or less charac~
teristic of the substan~e being reduced (reduction potential) .
• After the decomposition potential has been attained, the
current increases ra,pidly until a final (or saturation) value is
reached, the latter being related to the concentration of the
substance being reduced. If a llumqer of metalilc ions or
oxidising agents are present and the respective reduction
potentials differ by more than 0·2 volt, it is possible to deter-
pline the concentration of each in one experiment. This
forms the basis of the use of the polarograph in quantitative
analysis.
The instrument has been developed chiefly, by J. Heyrovsky
and his co-wotkers (1925 -+). Its use will.be evident from
Fig. 1-25. The solution to be analysed is contained in a small
A. =z mercury anode overlaid
by
E - the solution under test
, K = tbe dropping mercury
\ cathode
\ B _ Source of current (stQJ:'
'\\ W
age battery)
potentiometer wire coil
1'::=
\ ,, S = sliding contact
H =motor
PJ -= polarogram
G = moving coil reflecting
galvanometer
L = illuminating lamp
C -= condenser
FIG. 1-25. Recording the course of electrolysis with a droppillg mercury

cathode.
electrolytic vessel above a layer of mercury ~ which serves

as anode. The dropping mercury cathode is shown at K;
small drops of mercury, fed from an adjustable storage vessel,
are discharged slowly from a fine glass capillary into the
electrolyte. The surface of the dropping cathode is thus
constantly renewed, and hene,e the conditions under which
reduction takes place are readily reproducible. The poteritial
across the cell is slowly increased by means of a precision
rotating potentiometer W which is directly geared to the
recording drum P around which is wrapped a piece of photo~
graphic paper. The current flowing 'through the cell is
recorded on the drum py a spot of light from the mirror of the
galvanometer G, the sensitivity of which can be accurately
varied between 1 and 10,000 by means of a switch; the
galvanometer is connected in series with the cell through
the condenser C.
No current is passed by the cell until the deposition or
reduction'potential of one of the substances present in solution
is reached. A sharp increase. in current then sets in, i.e.,
the mercury-water interface becomes denuded of reducible
material by electrolysis, and iurther electrolysis is limited
by the rate at which fresh material diffuses to the interface
from the main bulk of_the solution. Further increase in the
voltage results in no further increase in current until the next
deposItion potential is reached, when the phenomenon is
repeated. In this way a curve '(polarogram) consisting of a
series -of steps is obtained: the voltage at which a step occurs
indicates qualitatively the $ubstance responsible and the
height of the step (under controlled conditions) its concentra-
tion. A typical polarogram (recorded twice to indicate its
repr~ducibility) is shown in Fig. 1-26. .
o 2 4 0 8 10 12 14 lb 18 20 2 x 10 - 'v It
I
•0 4 b 8 10 12 1.. 16 18 .,,'0 -1 Volt
FIG. 1-26. Polarogram of a solution of various metallic chlorides.
Determinations can be made in faintly acid (PH <j:: 3'5).

neutral or' alkaline solutions. With the exception of magne-
sium, all the common cations may be detected and determined
in this way. When two or more substances with the same or
but slightly different decomposition potentials are present
together, separation may be achieved by the addition of a
complex-forming reagent, such as. potassium cyanide or
ammonium oxalate; the resulting complex'ions frequently
The Theoretical Basis of Q1-(antitative Analysis 135
have deposition potentials differing from one another and
from those of the free ions. The method may also be applied
to the determination of chromates, iodates, bromates, chlor-
ates, nitrates, nitrites, dissolved oxygen, dissolved sulphur
dioxide, dissolved nitric oxide, hydrogen peroxide, organic
nitro compounds, azo compounds, aldehydes, ketones, un-
saturated acids, etc. For further details, the reader is
referred to the literature on the subject. * The great ~dis
vantage of this apparatus is that its cost is very high.
H. Methods based upon the use of X-rays, ele6tron diffrac-
tion, Raman and absorption spectra. The reader is referred
to the summary in the Annual Reports of the Chemical Society,
1939,35, 381-398.
* See, especially, reviews by J. Heyrovsky in W. Bottger, Physikalischl'
Methoden det' chemischen Analyse, Vol. II (Akad. Verlag. Leipzig, 1938) and
by 1. M. Kolthoff and J. 1. Lingane, The Fundamental Principles anI! Applica-
tions of Electrolysis with the Dropping Mercury Electrode"7tnd Heyrovsky's
Polarographic Method of Analysis (Chemical Reviews,. 1939, 24, 1-95).
W. Edwards and Co., of Vaughan Road, London, S.E.5, England, market
an excellent model of the polarograph, and issue a pamphlet which contains
much valuable information and many literature references. Figs. 1-25 and
1-26 are reproduced by courtesy of this firm.
An inexpensive apparatus'incorporating the dropping mercury electrode
system (the Fisher " electrodropode ") is marketed by the Fisher Scientific
Co., of Pittsburgh, U.S.A.
THEORY OF GRAVIMETRIC ANALYSIS
I, 54. Gravimetric anaiysis.-Gravimetric analysis o~
quantitative analysis by weight is the process of isolating and
weighing an element or a definite compound of the element in
as pure a form as possible. The element or compound is
separated from a weighed portion of the substance being
examined. A large proportion of the determinations of gravi-
metric' analysis is concerned with the transformation of the
element or r~dical to be determined into a pure stable-com-
ponnd, which can be readily converted into a form suitable
for weighing. The weight of the element or radical may then
be readily calculated from a knowledge of the formula of the
compound and the atomic weights of the constituent elements.
The separation of the element or 9f tlie compound contain- .
ing it may b~ effe.cted in a number of ways, the most important
of which are: (a) precipi~ation methods, (b) volatilisation or
evolution: methods, (c) electro-analytical methods, and (d)
miscellaneous physical methods. It is proposed to utilise
these four as the basis for the classification of gravimetric
methods, and to discuss the theory of each in turn.
It'may be mentioned at this stage that the weat advantage
of gravimetric over volumetric analysis is that the constituent
may be seen and examined for the presence of iplpurities and
a correction applied, if necessary; their disadvantage is that
they are generally more time-consuming.
1,55. PRECIPITATION METHODS
These are perhaps the most important with which we are
concerned in gravimetric analysis. The constit~ent .being
determined is precipitated from solution in a form which is
so slightly soluble that no appreciable loss occurs when the
precipitate is sepilrated by filtration and weighed. Thus in
the determination of silver, a solution of the substance is
treated with an excess of sodium or potassium chloride
solution, the precipitate is filtered off, well washed to remove
soluble salts, dried at 130-150°C, and weighed as silver
chloride. Frequently the constituent being estimated is
weighed iIi a form other than that in which it was precipi-
. tated. Thus magnesium is precipitated as magnesium
ammonium phosphate Mg(NH 4 )P0 4 , 6H 20, but is weighed,
after ignition, as the pyrophosphate Mg2P20" The following
Table contains the forms in which the compounds of ·the
common elements are usually weighed: this includes the
electro-analytical methods discussed in Section 1, 64 et seq.
136
The Theoretical Basis oj Q'uantitative Analysis 137
TABLE XVII. FORMS IN WHICH ELEMENTS AND RADICALS ARE CoMMONLY
P~ECIPITATED AND WEIGHED IN GRAVIMETRIC ANALYSIS.
Element Precipitated as Weigh;d as ~

------.------------------------
Pb PbSO.;
I. 2. PbCrO,;
__----------.-----------
PbSO,; 2. 'PbCrO, ;
I.
3· PbMoO,; 4. PbO,. 3· PbMoO,; 4· PbO •. •
--·-11------------,- - - - - - - --------
Ag I. Agel; 2. Ag. I. AgCI ; 2. Ag.·
Hg I.HgS; 2. Hg zinc thio- I. HgS; 2. Hg[Zn(CNS).);

cyanate; 3. Cu etbylenedia- 3. [Cu en.} [HgIJ.
mine mercuri-iodide.
Bi I. BiOI; 2. Bi pyrogallate; I. BiOI; 2. Bi(CaH.O.l ;

3. B~ ..cup.ferro~, complex; 3· Bi.O.; 4. Bi(C.H,ONl.·
4. BI. oxmate, H.O.
Cd I. CdMoO.; 2. Cd .8-naph- I. CdMoO,;
tha-quinoline complex; 3. Cd 2. [(ClSH.N).H.] (CdI,) ;
quinaldinate; 4. Cd pyri- 3. Cd(Cl~H.NO.l.;
dine thiocyanate; 5. Cd. . [' 4. [Cd(C.H.N).](CNS). ; 5. Cd.·
C'Il' I. Cu.(CNS).; 2. Cu ben- I. Cul(CNS). ;
zoinoxime; 3. Cu salic:9'lal- 2. Cu(CuHuO.N);
doxime; 4. Cu quinaldinate; . 3. Cu(C,H.O.N).;
5. Cu pyridine thiocyanate; 4. Cu(C 10H,NO.).. HIO;
6. Cu ethylenediamine mer- 5. [Cu(C,H.N).j(CNS).;
curi-iodide; 7. Cu. 6. [Cuen.] [Hg)IJ; 7. Cu.·
As I.1\5.5.; 2. Mg(NH.lAsO.,
6H.0; 3. NH,UO.AsO"
xH.O.
Sb I. Sb.S. ;
2. Sb pyrogallate.
Sn I. SnO., xH.O; 2. Sn cup- i. SnO,; 2. SnO,; 3.SnO•.

ferron complex; 3. Sn phenyl- I
arsonate-.
Mo. I. PbMoO,; 2. Mo I. PbMoO.; 2. MoO. ;.

benzoinoxim,e; 3. Mo .. oxin- 3· MoO.(C.H.ON)z .•
ate.-'
Se I.5e. I.5e.
Te I. Te; 2. TeO,. I.Te; 2.Te.
Pt 1. Pt; 2. Pt.
Pd Pd(CN).; 2. Pd dimethyl-
I.
g1yoxime; 3. Pd a-nitroso-p-
naphtholate ; 4. Pd a-nitro-p-
naphtholate.
1 - - - - 1 - - - - - - - - - - - - - - -1_ - - - - - - - - - -
Au I. Au. I
I. Au.
• Those elements (and PbO.) marked with an asterisk are electrolytic

determinations.
F*
TABLE XVlI---continued .
Enment
. Precipitated as W~ighed as
Al . I.Al(OR).; z. Basic Al suc- I. AI,O.; z. AI.O.;
cinate; 3. Al" oxinate." 3· Al(C.H.oN) 3'
Be I. Be(OHJi. I. BeO.
Cr I. Cr(OH)a: z. Rg.CrO.; I.Cr,Os; z. CraO. ; 3,
3. BaCrO., BaCrO•.
Fe 1:. Fe(OH).; z, Fe cupferron I. FeaO a ; z. Fe.O a,
complex.
Ni
. I.Ni dirnethylglyoxime; z. Ni
salicylaldoxime; 3. Ni pyri-
dine thiocyanate; 4. Ni 11-
I. Ni(C,H,O.N.).;
z. Ni(C,H,O.N).;
3. [Ni(C,H,N).](CNSh;
benzil-dioxime; 5. Ni. 4· Ni(C 14H 11 0.N.).; 5. Ni.*
Co I. Co a-nitroso-,B-naphtholate ; I. Co(C 1oH.O.N) •• 2H.O;
z. Co tt-nitro-,B-naphtholate; z. Co(C1oH.o.N)a; 3.Co,O,
3. Co phenylthiohydantoate (impure) and then as I. z. 4 or
complex; 4. Co pyridine thio- 5; 4. [Co(C&H&N),] (CNS). ;
.cyanate ; • 5. Co. 5. Co.·
Zn
. I.
2.
Zn(NH,)PO •• 6H 20 ;
ZnS; 3. Zn quinaldinate;
I.
2.
Zn(NH.)PO. or as Zn.P,O,;
ZnS or as I;
4· Zn pyridine ~hiocyanate; 3. Zn(C;: oH.O,N) •• 2H.O;
5. Zn. 4:[Zn(C,H.N).](CNS).; 5. Zn.'"
Mn I. Mn(NH.):f~O•• H.O. I.Mn,P,O,
or as Mn(NH.lPO •. H 2O.
V I. Hg.VO.; 2. Ag.VO,. I. V,05; 2. Ag.VO,.
U (UO.) I. (NH.J.U.O, ;
. I. U,O.;
2. U " oxinate." 2. UO,(C.H.ON).. C.HO,N.
Th I. Th(C,O.), ; 2. Th sebacate. I. ThO.; 2. ThO •.
Ce I. Ce(IO.),. I. CeO,.
Ti I.Ti" oxinate" ; I. TiO(C.H,ON). ; 2. TiO, ;
2.TiO(OH). with tannin and 3. TiO._; 4 TiOt ·
antipyrine; 3. TiO(OH) •.
with selenious acid; 4. Ti
para-hydroxy-phenylarsonate.
Zr I. ZrH,(PO.l. ; 2. basic Zr I. Zr.P.O r ; 2. ZrO. ;

selenite; 3· Zr n-propylar- 3· ZrO,; 4. ZrO,.
sonate; 4. Zr para-hydroxy-
phenylarsonate.
Tl I. Tl.CrO,; 2. Tl.[Co(NO.).l. I. TI.CrO.; 2. TI.[Co(NO.).l.

Ca I. CaCOO.,H.O ; 2. CaCO. ; .I.CaC.O•• H.o ;. 2. CaCO. ;
3· CaO . 3·CaO.
TABLE XVII-continued
.!lement Precipitated as Weigh_ed_as_ _ _ _ _
Sr I. srSo,;SrC.O"H.O.
3. r. SrSO•. ; 2. srO .
Ba I. BaSO,; 2. BaCrO •. .l:_BaSO.; 2. BaCrO•.
-ru:g- I. Mg" oxinate " ; r. Mg(C.H.ONl. ;
2. Mg(NH.)PO,,6H.o. 3. Mg(NH,)P0.,6H.0 or as
• Mg.P.O •.
~ I. Na.SO,; 2. Na zinc uranyl
acetate; 3. Na magnesium
uranyl ace1ate.
K -I. K.[PtCI.J; 2. KCIO.; I. K.[PtCI,]; 2. KCIO.;

3· KISO.; 4. K.Na[Co(NO.).], 3. K.SO.; 4. K.Na[Co(NO.).],
H.O. ' H.O.
Li I. Li.SO,; 2 .. by extraction as I. LizSO,; 2. LiCl.
LiC!.
(NH,).fPtCl.J. _____ _
I.
r. ,Benzidine tungstate; WO •.; 2. WO. :
I.
2. Tannin-antipyrine-tungstate 3. WO.(C.H.oN).
complex; 3. Tungstyl .. oxin-
ate."
Anions
CCBr-! AgCl, AgBr. AgCI, AgBr.
-1-' -1--:-----
I.
I. AgI; 2. PdI,.
1
I.
I. AgI; 2. PdI •.
I
CNS-
I r. AgCNS;
------1
2. CUlleNS).; I. AgCNS ; 2. Cul(CNS).:.
3· BaSO,. 3· BaSO,.
I
1------------------------1-------------------
~~I_I_._A_g_C_N_·__________ I-I-.-A-g-C-N-·--------------1
F- I_I_._P_b_C_IF_;_2_.C_a_F_z~;
3. (C.H.).SnF. ,
_____ I-I-.-P-b-C-I-F-;-2-.Ca_F
3· (C.H.l.SnF. __z_;______ 1
CIO ...- I. AgCl. I.AgCl.
CIO.- I. AgCI; 2. KCIO,. I. AgCI; 2. KCIO•.
SO,--,
S--, I. BaSO,. -: .. BaSO•.
SO.--,
S.O.--
PO,-- Mg(NH.lPO.,6H.o ;
I.
~.
(NH,l.PO.,12MoO.,,,
2HN0 3,H.o
I~~~==~I---~~~-------------
HPO. r. Hg,CI. ;
H.PO,- 2. Mg(NH.lPO,,6Hp.
1----1--
C.O.-- I. CaCP.,H'.O. I. CaC.O.,H.Oor.CaCO.or,CaO.
SiO.--, I. SiO.,xH.O. I. SiO•.
SiF.--
t-------------~L-----I-------------------
NO.- I I. Nitron nitrate.
"The factors which determine a successful analysis by pre-
cipitation are:
(A) The precipitate must be so sparingly soluble that the
precipitation may be made quantitative. In practice this
usually means that the quantity remaining in solution does
not exceed the minimum detectable by the ordinary analytical
balance, viz., 0·1 mg. . .
(B) The physical nature of the precipitate must be such
that it can be readily separated from the solution by -filtration,
and can be washed free of soluble impurities. These condi-
tions require that the particles are. of such size that they do
not pass through the filtering medium, and that tqe particle
size is unaffected (or, at least, not diminished) by the washing
process.
(C;) The precipit<;lte must be convertible into a pure sub-
stance of definite chemical composition; this may be effected
either by ignition or by a simple chemical operation, such as
evaporation, with a suitable liquid.
Factor (A), which is concerned with the completeness of
precipitation has already been dealt with in connexion with
the solubility product principle (Sections I, 8 and 9). Therein
are·discuss.ed the influence upon the solubility of the precipi-
tate of a salt with a common ion, of salts with no common ion,
of acids and bases, and o( temperature.
I t was assumed throughout that the compound which
separated out from the solution was chemically pure, but
this is not always the case. The ,Purity of the preGipitate;
depends inter alia upon the substance~ present in s9lution
both before and after the addition of the reagent, and also
upon the exact experimental conditions of precipitation. In
order to understand the influence of these and other factors,
it will be necessary to give a short account of the properties
of colloids.
I,56. The colloidal state.-The colloidal state of matter is
distinguished by a certain range of particle size, as a conse-
quence of which certain characteristic properties become
apparent. Before discussing these, mention must be made of
the various units which are employed in expressing small
quantities. The most important of these are:
Ip, :__ 10- 3 mm. ,Imp, = Ip,p, = 10- 6 mm.
1 Angstrom unit = lA. = 10-10 metre = 1O-7- mm. = 0·1 mf'.

Colloidal prop~rties are, 4I general, exhibited by substances
of particle size ranging between 0·1 p. and 1 mp.. Ordinary
quantitative filter paper will retain particles up to a diameter
of about 10- 2 mm. or 10ft, so that colloidal solutions in this
respect behave like true solutions (size of molecules is of the
order of 0·1 mft or 10-8 cm.). The limit of vision under the
microsc9pe is about 0'21-'. If a powerful beam of light is
passed through a colloidal solution and the solution is viewed
at right angles to the incide~t light, a scattering of light is
observed. This is the so-called Tyndall effect. True'solu-
tions, i.e., those with particles of molecular dimensions, do
not exhibit a Tyndall effect, and are said to be " optically
empty.'! Use is made of the Tyndall effect in the ultra-
microscope; here the Tyndall cone or beam is observed in a
microscope which. is situated at right angles to the path of the
incident light., The diffraction images are thus seen, and it is
possible to observe the -light scattered by each particle separ-
ately. The limit of visibility under the ultra-microscope is
about 10 mp.. .
By the use of X-rays the nature of the smallest unit of
colloidal substances may be ascertained. It has been found
that most substances consist of minute crystalline particles;
a few, such ,as silica and stannic oxide are amorphous. An
intermediate stage is also possible: a gradual development
of crystalline particles may- occur with some amorphous
substances upon" ageing" or with suitable treatment, such as
hot digestion with water or electrolytic solutions (Section
1, 58). .
An important consequence of the smallness of the size of
the particles is that the ratio of surface to weight is extremely
large. Phenomena which depend upon the size of the surface,
such as adsorption, will therefore play an important part
with substances in the colloidal state. Table XVIII clearly
shows the influence of particle size in connexion with a one
centimetre cube decimally divided.
The characteristic properties of colloidal particles are:-
(a) Th,ey exhibit a Tyndall effect when viewed with proper
illumination.
(b) They may be separated from true solutions of substances
by means of a collodion or parchment membrane, i.e., by
the process of dialysis.
(c) They may be regarded as possessing, electrical charges
since they migrate under the influence of a suitable potential
gradient. ' .
TABLE XVIII.INCREASE IN NUMBER AND TOTAL SURFACE OF PARTICLES AS A
ONE CENTIMETRE CUBE IS DECIMALLY DIVIDED.
Number of particles L81Igth of edge in em.) Total surface in sq. em.
1 1 6
10· 10-2 6 x 10'
IOu 10-4 (= lp.) 6 X 104
101• 10-· 6 X 105
1021 10-' (= Imp.) 6 X 10' (= ]·48 acres)
IOu 10-8 (= 11..) 6 x 10 8
~ .
For convenience, we may divide colloids into two main
groups, designated as suspensoids and emulsoids. * The
chief properties of each class are summarised jn the following
table, although, it must be emphasised, that the distinction
is not an absolute one, since some gelatinous precipitates (e.g.,
aluminium and other metallic hydroxides) have properties
intermediate between those of suspensoids and emulsoids.
Suspensoids _ (or Lyophobic

Colloids)
I Emulsoids (or Lyophilic
Colloids)
1. The solutions (or sols) are 1. The solutions are very vis-
only slightly viscous. Exam- cous; they set to jelly-like
pies: sols of. metals, silver masses known a~ gels. Ex-
halides, metallic sulphides, etc. amples ~ sols of silicic acid,
stannic oxide, g~atin.
2. A comparatively minute 2. Comparatively large con-
concentration of an electrolyte centrations of electrolytes are
results in flocculation. The required to.cause precipitation
change is, in general, irrever- f'salting out "). The change
sible; water has no effect is, in general, reversible, and is
upon the flocculant. effected by the addition of a
solvent (water).
3. Suspensoids, ordinarily, 3. Most· emulsoids change
have an electric charge of their charge readily, e.g., they
definite sign, which can be are positively charged in acid
changed 0 n I y b Y s P e cia 1 medium and negatively charged·
methods. in an alkaline medium.
4. The ultra-microscope re- 4. Only a diffuse light COne is
veals bright particles in vigor- exhibited under the ultra-
ous motion (Brownian move- microscope.
ment).
• For a detailed account, see H. Kruyt, Colloids, 1930 (Chapman and Hall).
The process of dispersing a gel or a flocculated solid to form a
sol is called peptisation.
The stability of colloidal solutions is intimately associated
with the electrical charge on the particles. Thus in the forma-
tion of an arsenious sulphide sol by precipitation with hydro-
gen sulphide in acid solution, sulphide ions are primarily
adsorbed (since ~very precipitate has a tendency to adsorb
its own ions), and in order to maintain the electro-Reutrality
of .the solution, an equivalent quantity of hydrogen ions is
secondarily adsorbed. The hydrogen ions of. other ions
which are secondarily absorbed, have been termed counter
ions. Thus the so-called electrical
double layer is set up between
the particles and the solution. An
arsenious sulphide particle is repre-
sented diagrammatically in Fig.
1-27. The colloidal particle of
arsenious sulphide has a negatively
charged surface, with positively
charged counter ions which impart H+'H+
a positive charge to the liquid F 97
, immediately surrounding it. If afi !G. 1-_ •
electric current is passed through the solution, the negative
particles will move towards the anode, although by virtue of
the fact that the colloidal particle consists of an aggregate of
arsenious sulphide molecules, its mobility will be less than
that of ordinary anions.
If the electrical double layer is destroyed, the sol is no longer
stable, and the particles will flocculate since the concentration
is in excess of the solubility px;oduct. Thus if. barium
chloride solution is added, parium ions are preferentially
adsorbed by the particles; the orientation of the surface is
disturbed and the charge disappears. After flocculati_on, it
is found that the dispersion medium is acid owing to the
liberation of the hydrogen counter ions.' It appears that-ions
of opposite charge to those primarily adsorbed on the surface
are necessary for coagu1ation. The minimum amount of
electrolyte necessary to cause flocculatibn of the colloid is
called jhe flocculation or coagulation' value. It has been
found that the latter depends primarily upon the valency of
the ions of the opposite charge to that on the colloidal
particles: the nature of the ions has some influence
alse This is clearly shown by the results collected in Tabl~
J
XIX.
144 Quantitative' Inorganic Analysis
TABLE XIX. COAGULATION VALUES IN MILLI-MOLS OF COAGULATING ION
PER LITRE.
NEGATIVE ARSENIC TRISULPHIDE POSITIVE HYDRATED FERRIC

SOL OXIDE SOL I
Salt Coag. Value Salt Coag. Value
AICI •• . 0·062 K.Fe(CNl, 0·067 .

AI.(SO.l. 0·074 K.Fe(CN), 0·096
FeCI.
CaCl.
... 0'136
0'649
K.SO.
K.Cr,O.
0'22
0·19
BaCI. 0·691 K.CrO. 0'33
MgCl. 0·717 K.C.O, 0·24
I3a(NO.l. 0·687 KBrO. 31
KCI 49'5 KCNS 47
NaCl I 51·0 KCl 103
LiC} .... 58'4
50·0
KNO.
KBr . 131
138
KNO,
HC}
. 30·8 KI 154
If two sols of opposite sign are !flixed, mutual coagulation

usually occurs owing to the neutralisation. of charges. The
above remarks apply largely to sllspensoids. Emulsoids are
generally very more difficult to coagulate than suspensoids.
If an emulsoid sol, e.g., of gelatin, is added to \a suspensoid
sol, e.g., of gold, then the suspensoid sol appears to be strongly
protected against the flocculating action of electrolytes. It
is probable that the particles of the emulsoid sol are adsorbed
by the suspensoid and impart their own properties to the
latter. The emulsoid is known as a protective colloid. This
explains the relative stability of the otherwise unstable gold
sols produced by the addition of a little gelatin. For this
reason also, organic matter, which might form a protective
colloi9, is generally destroyed before proceeding with an
inorganic analysis.
During the flocculation of a colloid by an electrolyte, the
ion of opposite sign to that of the colloid is adsorbed to a
varying degree on the surface; the higher the valenc~ of the
ion, the more strongly is it adsorbed. In all cases, the pre-
cipitate will be contaminated by sutface adsorption. Upon
washing the precipitate with water, part of the adsorbed
electrolyte is removed, and a new difficulty may arise. The
electrolyte concentration in the supernatant liquid may fall
below the coagulation value, and the precipitate will .pass
into colloidal solution again. This :phenomenon, which is
The Theoretical Basis of Quantitafive Analysis 145
known as peptisation, is of great importance in quantitative
analysis. By way of illustration, let us consider the precipi-
tation of silver by excess of chloride ions in acid solution and
the subsequent washing of the coagulated silver chloride with
water; the adsorbed hydrogen ions will be removed py the
washing process and::} portion of the precipi.tate will pass
through the filter. If, however, washing i~ carried out with
dilute nitric acid, no peptisation occurs. Fot this reason.
precipitates are always washed with a suitable solution of an
electrolyte which does not interfere with the subsequent steps
in the estimation.
The adsorptive properties. of colloids find a number of
applications in analysis, e.g., in the removal of phosphates by
hydrated stannic oxide in the presence of nitric acid, in the
use of adsorption indicators (Section I. 42), in the qualitative
detection and colorimetric estimation of elements and radicals
with many organic reagents (for example, aluminium with
" aluminon " Section V, 15).
1,57. Supersaturation and precipitate formation.-The
solubility of a substance at any given temperature in a given
solvent is the amount of the substance dissolved by a known
weight of that solvent when the substance is in equilibrium
with the solvent. The solubility depends upon the particle
size, when these are smaller than about 0·01 mm. in diameter;
the solubility increases greatly the smaller the particles owing
to the increasing role played by surface effects (compare
Table XVIII). The definition of solubility given above refers
to particles larger than 0·01 mm. A supersaturated solution
is one that contains a greater concentration of solute than
corresponds to the equilibrium solubility at the temperature
under consideration. Supersaturation is therefore an unstable
state whicl:;t may.be brought to a state of stable equilibrium
by the addition of a crystal of the solute (" seeding" the
solution) or of some other substance, or by mechanical means
such as shaking or stirring. The difficulty of precipitation of
magnesium ammonium phosphate will at once come to mind
as an example of supersaturation.
According to von Weimarn (1913, 1925) supersaturation
plays an important part in determining the particle size' of a
precipitate. He has deduced that the initial velocity of
precipitation is proportional to (Q - 5)/5, where Q is the
total concentration of the· substance that is to precipitate,
and 5 is the equilibrium solubility; (Q - 5) will denote the
146 Quantitative. Inorganic Analysis
supersaturation at the moment precipitation commences.
The expression applies approximately only when· Q is large
as ,compared with 5. TIfe influence of the degree of super-
saturation is well illustrat~ by von Weimarn's results for the
formation of barium sulphate Irom solutions of barium thio-
cyanate and manganese sulphate respectively. These are
collected in Table xx below. The results clearly show that
TABLE 'KX. SEPARATION OF BaSO, AT VARIOUS DEGREES OF SUPERSATURA-
TION (VON WEIMARN).
Concenlrailon Type of Precipitate

of Re,!gents (Q- 5)/5
7N 175,000 A gelatinous precipitate is formed, and

practically the whole of the water is ad-
sorbed; the containing vessel can be
inverted without the contents running
out. The gel is unstable, and growth of'
the large crystals at the expense of small
ones is very rapid; after a few hours the
precipitate becomes opaque. .
3N 75,000 'Gelatinous films formed; become turbid
after one minute.
N 25,000 Primary precipitate is curdy and of
j:olloidal dimensions. Particles appear as
points at a magnification'of 1500 x.
6'05N 1300 Primary precipitate consists of feathery
and star-shaped crystal skeletciJDs.
0'005N 125 Precipitate consists of compatt crystal
skeletons.
ca.O·OOlN 25 Solution· becomes opalescent during first
five minutes, and precipitation continues
, for 2 to 3 hours. After tl}at time crystals
have a length of 0·005 mm. -
ca.0·0002N 5 Precipitate appears after about a month.
At the end of six months, the length of
, the largest crystals is about 0·03 mm: and
their breadth 0·015 mm.
the particle size of a precipitate decreases with, inq-easing

concentration of the reactants. For the production .of a
crystalline precipitate, for which the adsorption errors will
be least and filtration will be easiest, (Q - 5) /5 should be as
small as possible. There is- obviously a practical limit to
reducing (Q - 5)/5 by making Q very small, since for a
precipitation to be of value in analysis, it must be complete
ih a comparatively short time and the volumes of solutions
involved must not be too large. There is, however, another
method which may be used, viz., that of increasing 5. For
example, barium sulphate is about fifty times' more soluble
in 2N hydrochloric acid: if O·lN solutions of barium chloride
The Theoretical Basis of Quant#ative Analysis 14:7
and sulphuric acid are prepared in 2N boiling hydrochloric
acid and the solutions mixed, a typical crystalline precipitate
of barium sulphate is slowly formed.
The applications of the above conceptions· are to be found
in the following recognised procedures in gravimetric analysis:
1. Precipitation is usually ~arried out in hot solutions. since
the solubility generally increases with temperature.
2. Precipitation is effected in dilute solution and the
reagent is added slowly and with thorough stirring. The
slow addition results in the first particles precipitated acting
as nuclei for the -deposition of new particles as they are
formed.
3. A suitable reagent is often added to increase the solu-
bility of the precipitate and thus lead to larger primary
particles.
I, 58. The purity of the precipitate. Coprecipitation.-
When a precipitate separates from a solution, it is not always
perfectly pure: it may contain varying amounts of impurities
dependent upon the. nature of the precipitate and the con-
ditions of precipitation. The contamination of the precipi-
tate by substances which are normally soluble in the mot!ter
liquor is termed coprecipitation. We must distinguish be-
tween two important types of coprecipitation. The first is
concerned with adsorption at the surface of the particles
exposed to the solution. and the second I'elates to the occlusion
of foreign substances during the process of crystal growth from
the primary particles. .
With regard to surface adsorption, this will, in general, be
greatest for gelatinous precipitates and least for those of
pronoun,ced macro-crystalline character. Precipitates with
ionic lattices appear to conform to the Paneth-Fajans-Hahn
adsorption rule, which states that" those ions whose com-
.pounds with the oppositely charged constituent of the lattice
are slightly soluble in the solution in question are well ad-
sorbed by the ionic lattice." Thus silver iodide adsorbs
silver acetate much more strongly than silver nitrate under
comparable conditions since tne former is the less soluble.
The deformability of the adsorbed ions and the electrolytic
dissociation of the adsorbed compound also have a consider-
able influence; the smaller the dissociation of, the compound,
the greater is the adsorption. Thus hydrogen sulphide, a
* For a detailed discussion, see T. B. Smith. AnalytiGal ProGessIs. 1929
(Arnold and Co.).
weak electrolyte, is strongly adsorbed by metallic sulphides.
The second type of coprecipitation may be visualised as
occurring during the building up of the precipitate from the
primary particles. The latter witl be subject to a certain
amount of surface adsorption, and during their coalescence
the impurities will either be partially eliminated if .large
single crystals are formed al\d the process takes place slowly,
or, if coalescence is rapid, large crystals composed of loosely
bound small crystals may be produced and some of .the
impurities may be entrained within the walls of the large
crystals. If the impurity is isomorphous or forms a solid
solution with the precipitate, the amount of coprecipitation
may be very large, since there will be flo tendency for elimina-
tion during the" ageing" process. The btter actually occurs
during the precipitation of barium sulphate in the ,presence
of alkali nitrates; in this particular case X-ray studies have
shown that the abnormally large coprecipitation (which may
be as high as 3·5 per cent if precipitation occurs in the presence
of high concentrations of nitrate) is due to the formation of
solid crystals. Fortunately, however, §uch cases are com-
paratively rare in analysis ..
Appreciable errors may also be introduced by post-precipi-
tation. This is the precipitation which occurs on the surface
of the first ·precipitate after its formation. ;!t occurs with
- sparingly soluble substances which form supersaturated
solutions; they usually have au ion in common with the
primary precipitate. Thus in the precipitation of calcium
as oxalate in the presence of magnesium, magnesium oxalate
separates out gradually upon the calcium oxalate; the longer
the precipitate is allowed to stand in ~ontact with the solution,
the greater is the error due to this cause. A similar effect is
observed in the precipitation of copper or mercuric sulphide
in O·3N hydrochloric acid in the presence of zinc ions; zinc
sulphide is slowly post-precipitated.
It is convenient to consider now the influence of digestion.
This is usually carried out by allowing the ptecipitate to
stand for 12 to 24 hours at Foom temper.ature, or sometimes
by warming the precipitate for some time, in contact with the
liquid from which it was formed: the object is, of course, to
obtain complete precipitation in a form which can be readily
filtered. During the process of digestion or of the ageing of
precipitates, at least two changes occur. The very small
particles, which have a greater solubility than the larger ones,
will, after precipitation has occurred, tend to pass into solu-
The Theoretical Basis of Quantitf!tive Analysis 149
tion, and will ultimately redeposit upon the larger particles ;,
.the coprecipitation on the minute particles is thus eliminated
and the total coprecipitation on the ultimate precipitate
reduced. The rapidly formed crystals are probably of
irregular shape and possess -a comparatively large internal
surface; upon digestiori these tend to become more regular
in character, thus resulting in a decrease in the area of the
internai surface and a consequent reduction of adsorption.
The net result of digestion is usually to reduce the extent of
coprecipitation and to increase the size of the particles,
rendering filtration easier.
I, 59. Conditions of precipitation.-No universal rules for
precipitation can be given which are.applicable to all cases of
precipitation, but, with the aid of an intelligent application
'of the facts enumerated in the foregoing paragraphs, a number
of fairly general rules may be st~ted : •
1. Precipitation should be carried out in dilute solution.
due regard being paid to the solubility of the precipitate, the
time required for filtration "and the sub~equent operations to
be carried out with the'filtrate. This will minimise ,the errors
due to coprecipitation.
2. The reagents should be mixed slowly and with constant
stirring. This will keep the degree of supersaturation small
and will assist the growth of large crystals. A slight exCess
of t~e reagent is, all thai is generally required; in exceptional
cases a large excess may be necessary. In some instances the
order of mixing the reagents may be important. Precipita-
tion rimy "be effected under conditions which increase the
solubility of the precipitate, thus further reducing the degree
of supersaturation (compare Section I, 58).
3. Precipitation is effected in hof solutions, provided the
solubility and the stabilify of the precipitate permit. Either
one or both of the solutions should be heated to just below the
boiling point or other most favourable temperature. At the
higher temperature (a) the solubility is increased with a con-
sequent reduction in the degree of supersaturation, (b)
coagUlation is assisted and sol formation decreased, and (c)
the velocity of crystallisation is increased,' thus leading to
better-formed crystalsJ
4. Crystalline precipitates 'should be digested for as long as

practical, preferably overnight, except in those cases where
post-precipitation may occur. As a rule, digestion on the
steam bath is desirable. This process decreases the effect of
coprecipitation and gives more readily filterable precipitates.
Digestion has little effect upon amorphous or gelatinous
precipitates.
5. The precipitate should be washed with the appropriate
dilute solution of qn electrolyte. Pure water may tend to
cause peptisation. (For theory of washing, see Section I, 60
below.)
6. If the precipitate is still appreciably contaminated as a
result of coprecipitation or other causes, the error may often
be reduced by dissolving it in a suitable solvent and them
reprecipitating it. The amount"of foreign substance present
in the second precipitation will be small, and consequently
the amount of the entrainment of the precipitate will also be
small. e
I, 60. Washing of the precipitate.-the -experimental

aspect of this important.subject is dealt with in Section II, 34.
Only some general theoretical considerations will be given
here. Most precipitates are produced in the presence of one
or more soluble compounds, and i.'I: is the object of the washing
process to remove thes~ as q>mpletely as possible. It is
evident that only surface impurities will be- removed in this
way. The composition of the wash solution will depend
upon the solubility.and chemical properties of the precipitate
and upon its tendency to undergo peptisation, the impurities
to be rem0ved, and the influence of traces of the wash liquid
upon the subsequent treatment of the precipitate before
weighing. Water cannot, in general, be employed owing to
the possibility of producing partial peptisation of the precipi-
tate and, in many cases, the introduction of small losses as a
consequence of the slight solubility of the precipitate. A
solution of some electrolyte is employed. This should pos-
sess a common ion with the precipitate in order to ·reduce
solubility errors, and should easily be volatilised in the pre-
paration of the precipitate for weighing. For these reasons,
ammonium salts, ammonia solution and dilute acids are
commonly employed. If the filtrate is required in a subse-
quent determination, the selection is limited to substances
which will not ·interfere in the sequel. Also hydrolysable
substances will necessitate the use of_solutions containing a
salt which will depress the hydrolysis (compare Section
I, 18). As examples of washing liquids we may mention very
dilute nitric acid for the silver halides, "and ditute ammonium
oxalate solution for calcium oxalate. Whether. the wash
liquid)s employed hot "or at some otlier temperature will
depend primarily upon the solubility of the precipitate;
if permissible, hot solutions are to be preferred, because of the
greater solubility of the foreign substances and the increased
speed of filtr(!.tion.
Solubility losses are reduced by employing the minimum
quantity of wash solution consistent with the removal of
impurities. It can be readily shown that washing is more
efficiently carried out by the use of many small portions of
liquid than with a few large portions, the volume being the
same in both instances. Under ideal conditions, where the
foreign body is simply mechanicaIly associated with the
particles of the precipitate, the following expression may be
shown to hold * :
.X» = Xo ( ~: v )"
where Xo is the concentration of impurity before washing, x"

is the concentration of impurity after n washings, u is the
volume in ml. of the liquid remaining with the precipitate
after draining, and v is the volume in ml. of the solution used
in each washing. It follows from this expression that it is
best (a) to allow the liquid to drain as far as possible in order
to maintain u at a minimum, and (b) to use a relatively small
volume of liquid and to increase the number of. washings.
Thus if u = 1 ml. and v = 9 ml., five washings would reduce
the surface impurity to 10-6 of its original value; one washing
with the same volume of liquid, viz., 45 ml., would only reduce
the concentration to 1/46 or 2·2 X 10- 2 of its initial concen-
tration.
In practice the washing process is not quite so efficient as
the above simple theory would indicate, since the impurities
are not merely mechanically associated with the surface.
Furthermore, solubility losses are not so great as one would
expect from the solubility data because the wash solution
passing through the filter iS'11ot saturated with respect to the
precipitate. ,Frequent qualitative tests must be made upon
portions of the filtrate for some foreign ion which is known to
be present in the original solution; as soon as these tests are
negative, the washing is discontinued.
• T. B. Smith, Analytical Processes, 1929, p. 48 (Arnold and Co.); see also'

J. W. Mellor and H. V. Thompson, A Treatise on Quantitative Inorganic
Analysis, 1938, p. 84 (C. Griffin and Co. Ltd.).
152 Quantitative Inorganic A na!ysis
QUANTITATIVE SEPARATIONS BASED UPON PREOJPITA-
TION METHODS
I. 61. Fractional precipi~tion:-The simple theory of
fractional precipitation has been given in Section I. 11. It
was shown that when the solubility products of two sparingly
soluble salts having an ion in common differ sufficiently, then
one salt will precipitate almost comple~ely before the other
commences to separate. This separation is actually possible
for a mixture of chloride and iodide, but in other cases the
theoretical predictions must be verified experimentally
because of the danger of coprecipitation (Section I, 58)
affecting the results. Some separations based upon fraction,ai
precipitation, which are of practical importance, will now be
considered.
(A) Precipitation of sulphides. In order to fully under-
stand the separations dependent upon the sulphide ion, we
shall consider first the quantitative relationships involved in a
saturated solution of hydrogen sulphide. The following'-
equilibria are present:
HzS ~ H+ + HS-;
HS- ~ H+ + S--.
[H+] X [HS-] / [HzS] = Kl = 9·1 X lO-8 (i).
[H+] X [S--] I [HS-] = K2 = 1·2 X 10-15 (ii).
The very small value of K2 indicates that the secondary dis-

sociation and consequently [S--] is exceedingly small. It
follows therefore that only the primary ionisation is of im-
portance, and [H+] and [HS-] are practically equal in' value.
A saturated aqueous solution of hydrogen sulphide at 25°C, at
atmospheric pressure, is approximately 0·1 molar.. Sub-
stituting' [H+] = [HS-] and [HzS] = 0·1 in equation (i), we
obtain:
[H+] = "[HS-1 = v9'1 X 10 8 X 0'1 = 9·/jx 10-5 •
Both the equilibrium equations 'must be satisfied simul-

'taneously; by substitution of these values for [H+] and [HS-]
in equation (ii), we obtain:
9·5 X 10- 5 X [S--] = 1·2 X 10-15 X ~'5 X 10-5 ,
or [S--1 = 1·2 X 10- 1 ";
which is the value for K 2'

The Theoretical Basis of Quantit~tive Analysis '153
I~ we multiply equations (i) and (ii) together, we obtain:
[H+J2 . [S--) = 1.1 X 10- 22 (iji) ,
[HzS] .
or [S--] =,1,1 X 10-23 J [H+J2 (iii').
Thus the concentration of the sulphide ion is inversely pro-
portional to the square of the hydrogen ion concentration.
By varying the PH of the solution, the [S--] may be controlled;
in this way separations of metallic sulphides may be effected.
Let us first calculate the [S--] in a solution of 0·25 molar
hydrochloric acid saturated with hydrogen sulphide; this is
the concentration employed for the precipitation of the sul-
phides of the Group II metals in qualitative analysis. The
total concentration of H 2S will be approximately the same as
in aqueous solution, i.e., O·IM : the [H+] will be equal to that
of the completely dissociated hydrochloric acid, i.e., 0'25M,
but the [S-J will be reduced below 1·2 X 10-15 • Substitut-
ing in equations (i) and (ii) :
8 8
[ HS-] = 9·1 X 10- X [HzS]= 9'1 X iO- X 0-1 = -6 10- 8
[H+] 0-25 3 X ,
_ 1-2 X 10-15 X [HS-] 1-2 X 10-15 X 3·6 X 10-8
[S J = [H+] = 0'25
= 1·7 X 10- 22 •
Thus by changing the acidity from 9·5 X IO-5M (that present
in saturated hydrogen sulphide water) to 0-25M, the sul-
phide ion concentration is reduced from r'2 X 10- 15 to
1·7 X 10- 22 ,
With the aid of a table of solubility products of metallic
sulphides (see Section I, 8, Table IV), we can calculate whether
.certain sulphid~s will precipitate under any given conditions
of acidity and also the concentration of'the metallic ions
remaining in solution. Precipitation of a metallic sulphide
MS will occur when [M++] X [S--] exceeds the solubility •
product, and the concentration of metallic ions remaining in
the solution may be calculated from the ,equation:
[M++] _ SMS _ SMS X [H+P C)
- [5--] - 1·1 X 10- 22 X [HzS] IV ,
As an example we may consider the precipitation of _cupric

sulphide (Scus = 8·5 X 10-45) and ferrous sulphide
(SFeS = 1·5 X 10-18) from O'OIM solutions of the metallic
-l54' Quantitative Inorganic Analysis
ions in the presence of 0'25M hydrochloric acid. For cupric
sulphide, the solubility product is. readily exceeded
[S;-]=1'7x 10- 22 , [Cu++]=O·Ol and precipitation will occur
until [Cu++] = 8·5 X 10-45 X (H+P / 1·1 X 10- 22 X [H 2S],
= 8·5 X 10-45 X (0·25)2 / 1'1 X 10- 22 X 0,1,
= 5 X 10- 23 ,
i.e., precipitation is virtually complete. With ferrous sul-

phide, the solubility product cannot be exceeded ana pre-
cipitation will not occur under these conditions. If, however,
the acidity is sufficiently decreased, and consequently [S--]
increased, ferrous sulphide will be precipitated.
The case of zinc sulphide is of especial interest. Various
values are given in the literature for its solubility product:
the most trustworthy figures vary between 1 X 10- 24 and
8'X 10- 26 • If we accept the latter figure, then we should
expect precipitation to occur in a, say, O'OlM solution of zinc
ions in the presence of 0·25M hydrochloric acid, since the
S.P. should be exceeded; furthermore, the residual zinc ion
concentration should be 4·7 X 10-4 when computed as
described above. In practice, precipitation does not occur
at this acidity. This may be partly due to the great ten<;lency
tpat zinc ·sulphide possesses to remain in supe~saturated
solution, but is perhaps be,st explained as follows. The ab.ove
figure for the solubility product refers .to a solution in equili-
brium with relatively large particles, whereas for precipitation
to OCCUl: it is neces'sary that j:he S.P. of the particles actually
formed should be exceeded. It may well be that under the
above experimental' conditions these are extremely small,
thus possessing 'a greater solubility (Section I. 56) and a
greater solubility product; precipitation will therefore not
take place. This view is supported by the fact that post:
precipitation of zinc sulphide (compare Section I. 58) will
occur upon the surface of other metallic sulphides, such as
those of copper and mercury. It is possiDle to precipitate
zinc in acid solution provided the experimental conditions are
very carefully controlled, e.g., when the pH -of the solution
lies between 2 and 3 and ammonium salts ate present as
coagulants: this is attained by the use of a buffer mixture
of formic acid and ammonium formate, and ammonium
sulphate or of chloroacetjc acid and sodium acetate. It is
probable that large particles of zinc sulphide are initially
formed under these conditions.
(B) Precipitation and separation of hydroxides at controlled
hydrogen ion concentration or pH. The underlying theory is
very similar to that just given for sulphides. Precipitation
will depend largely upon the solubility product of the metallic
hydroxides and the hydroxyl ion concentration (or upon the
hydrogen ion concentration, since pH + pOH = pK",-
Section I, 17) of the solution.
We have seen that the sulphide ion concentration of a
saturated aqueous solution of hydrogen sulphide may be
controlled within wide limits by suitably changing the con-
centration of hydrogen ions-a common ion-of the solution.
In a like manner the hydroxyl ion concentration of a solution
of a weak electrolyte, such as ammonium hydroxide
(Kb =.1·8 X 1O-5), may be regulated by the addition of a·
common ion, e.g., ammonium ions in the form of the com-
pletely dissociated ammonium chloride. The magnitude of
the effect is best illustrated by means of an example. In a
0·1 molar ammonia solution, (he degree of dissociation is given
(Section I, .4) approximately by :
a = V Kc' = v''71-:·8'--x-l'''''0'----=-5-x----=-O-=·I = 0·013.
Hence [OH-] = 0'0013, [NH,,+] = 0',()013, and [NH"OH] =
0·0987. Let us cal'1ulate th~' change in'the degree of disso-
ciation and of [OH-] upon the addition of 0·5 gram molecule
(26'75 g.) of ammonium chloride to one litre of the 0'1 molar
ammonia solution. If a' is the degree of dissociation in the
presence of the added ammonium chloride, then :
[OH-] = a'e = O·la', and [NH"OH] = (1 - a'}c = 0·1,
. "
since a' may be taken as negligibly small. The addition of
the completely dissociated ammonium chloride, will of neces-
sity decrease the INH,,+] derived from the base and increase
[NH,OH] since Kb is constant under all conditions. Now, as
a first approximation, [NH,+] = 0·5.
Substituting this value in the equation:
[NH,+] X [OH-] 0·5 X O'la'
- - - - = 1·8 X 10-6 ,
[NH,OH] - 0·1
a' = 3·6 X 1rr 5 and [OH-] = 3·6x 10-8 •
Thus the addition of half an equivalent weight of ammonium
chloride to a 0'1 molar solution.of ammonium hydroxide has
decreased the hydroxyl ion concentration from 0·0013 to
0'000'0036, or has changed pOH from 2·9 to 5·4, i.e., the pH
has changed from 11·1 to 8·6.
156 Quantitft,tive Inorganic Analysis
An immediat~ application of the use of the ammonium
hydroxi<!e-ammonium chloride mixture may be made to the
familiar example of the prevention of precipitation of mag-
nesium hydroxide (S.P. 1·5 X 10-11). We can first compute
the minimum hydroxyl ion concentration neces'sary to prevent
precipitation it'l, say, 0·1 molar magnesium solution.
_ _ JSMg(OHlI_Jl'5 X 10 11 = ,1'.22 X 1O-C;M.
[OH ] - '[Mg+'+] - 0.1
or POH = 4·9 and pH = 14'0 - 4'9 = 9·l.
If we employ an ammonium hydroxide solution which is 0·1
molar, the concentration of [NH,+] ion as ammonium chloride
or other ammonium salt necessary to prevent the precipitation
of magnesium hydroxide can be readily calculated as follows.
Substituting in the mass action equation:
[NH,+] X [OH-] = 1.8 10-5
[NH,OH] x,
[NH +] X 1·22 X 10-&
" . = 1·8 X 10-5
0'1 '
or [NH4+] = 1·48 X -f(f-lM.
This corresponds to an ammonium ,chloride conce~tration of
1·48 X 10-1 X 53·5 = 8·2 g. per litre.
We Will now consider the conditions necessary for the
practically complete precipitation .of magnesium hydroxide
from a O'IM solution of, say, magnesium chloride. A POH
slightly in excess of 4·9 (i.e., pH = 9·1) might fail to precipi-
tate the 'hydroxide owing to supersaturation. Let us suppose
the hydroxyl ion concentration is increased ten times, i.e., to
pOH 3·9 or pH 10'1, then, provided no supersaturation is
present: •
11
[M ++] = SMg(OHh = 1'5 X 10- = O'OOIM
g [OH-P (1·22x 10-')2 '
i.e., the concentration of the magnesium ions remaining in
solution is O'OOIM, or on~ per cent of the magnesium ions
would remain unprecipitated. If pOH is changed to 2·9 or
pH to 11·1, it can be'shown in a similar way that the concen-
tration of the magnesium iOIli left in solution is ca. 1 X 10-5 M,
so that the precipitation error is 0·1 per cent, a negligible
quantity. We may therefore say that magnesium is pre-
Cipitated quantitatively at a pH of 11·1.
Our knowledge of the solubility products of metallic hydro-
oxides is, however, not very precise, so that it is not alway~
possible to make exact, theoretical calculations . •The pH
values at which various hydroxides begin to precipitate from
dilute solution are collecte~ in the following table, which is
based upon the work of R. T. S. Britton (1925).
pH VALUES AT WHICH VARIOUS HYDROXIDES ARE PRECIPITATED.
I
pH Metal Ion pH Metal Ion
II .
. Sn++,Fe+++,Zr++++
I
3 7 Fe++
4 Th++++ 8 Co++, Ni++, Cd++
5 Al+++ 9 Ag+, Mn++, Hg++.
6 Zn++, Cu++, Cr+++ 11 Mg++
One of the easiest, and also a widely employed, separation

is that of ferric iron from magnesium. This can be readily
effected, as we have just seen, by the use of ammonium chlor-
ide-ammonium hydroxide buffer mixture. By proper adjust-
ment of the pH of the solution it is possible to separate ferric
iron and aluminium from zinc, manganese, nickel, cobalt,
and the alkaline earths. This regulation of pH is achieved
by the addition of the apprppriate buffer mixture. It is
evident from the above Tal?le that some \5eparation~ will be
much more difficult than others: thus the separation of
aluminium and chrorp.ium is an example of a difficult separa-
tion. The use of ammonia solution is not always satisfactory
as it tends to produce, particularly in the case of the trivalent
metals, gelatinous precipitates, which are difficult to filter
and wash, and also have a marked tendency to adsorb other
cations ptesent in splution. For this reason other reagents
are employed for the separation.of trivalent ions fr.om divalent
ions: a few exa.mples will be mentioned below. Reference
must, however, be made to the use <;>f the organic bases, such
as phenyl-hydrazine, aniline, pyridine, quinoline, .etc. Thus
phenyl-hydrazine is particularly' valuable for separating
aluminium from ferrous iron; only the former is precipitated
as the hydroxide. The other organic bases are useful in the
sepatation of the rare earths. .
• In the basic acetate method a buffer solution cOllsisting of
acetic acid w_ith sodium or ammonium acetate is employed.
The acetates of such weak bases as ferric iron are hydrolysed
in hot solution and are completely precipitated as basic
acetates; the acetates of the strong bases of the divalent
metals are much less hydrolysed and, if the pH is adequately
controlled, are not precipitat~d. Tpe method is quite satis-
factory for the separation of ferric iron from divalent metals
such as zinc, manganese, cobalt, nickel, and copper; it is
fairly satisfactory for ,mixtures of iron and aluminium, but
not so suitable for aluminium alone; it is inapplicable for
chromium.
Fe(CzHaOz)a + 2H 20 ? 2H.C2 H a0 2 + Fe(OHh·C~Ha02·
A modification, introduced by Kolthoff (1934), is .to use a
benzoic acid-ammonium benzoate bllffer mixture, when a
good separation of ferric iron, aluminium and chromium from
the divalent metals is claimed.
Suspensions of carbonates (e.g., of barium, lead, and cad-
mium) and of oxides (e:g., of zinc, magnesium and mercury)
have been used for the same purpose, since they may be
regarded as exerting a buffer action. Their applications are,
however, limited; mention may be made of the use of barium
or cadmium carbonates for the separation of the trivalent
metals of Group IlIA from the divalent metals of Group IIIB.
1, 62. Organic p,ecipitants.-A number of ~rganic com-

pounds are availal'>le ~ich combine with inorganic ions or
compounds to form sparingly soluble and often coloured
compounds. With cations, the metal is held either by prim-
ary valencies or by co-ordinate links, or by both. Tliese
compounds usually have high m'olecular weights 'so that a
small amount of the ions will yield a rela~ively large amount of
the precipitate. The ideal organic precipitant should be
specific in character, i.e., it should give a precipitate with
only one particular ion. In few cases, however, has this ideal
been attained; it is more usual to find that the organic
reagent will react with a group of ions, but frequently by a
rigorous control of the experimental conditions it is possible to
precipitate only one of the ions of the group. Sometimes tHe
precipitated organic compound may be weighed after dPying
at a suitable temperature; in other cases the composition is
not quite definite and the substance is converted by ignttion.
to the oxide of the metal; in a few instances, a volumetric
method is employed which utilises the quantJ'tatively pre-
cipitated organic complex. Examples of each of these will
be given below.
A. Dimetbylg]Yoxime. This reagent (I) was discovered by
L. Tschugaeff in 1905
CHa - C = NOH
I (I)
CHa - C = NOH
OH 0
I t
CHa;-C = N" /N
= C - CHa
I Ni I (II)
CHa - C = N/ 'N = C - CHa
.j. I
o OH
and was applied by O. Brunck in 1907 for the determination
of nickel in steel. I t gives a bright red precipitate II or
Ni(C,H 70 2N 2)2 with nickel salt solutions; precipitation is
usually carried out in ammoniacal solution or in a buffer
'solution containing ammonium acetate and acetic 'acid. The
complex is weighed after drying at 1l0-120° C. Palladium
solutions also give a characteristic precipitate of the composi.
tion Pd(C,H7 0 2N 2h; this precipitation is carried 'out in cold
solutions containing 2 to 5 per cent by volume of hydrochloric
acid. If iron.and/or aluminium are present, citric or tartaric
acid must be added in order to prevent hydrolysis.
The reagent itself is insoluble in water, and is employed as
an approximately 1 per cent solulion in alcohol.
a-Furildioxime (III) has also been proposed for the deter-
inination of nickel; it gives a red
(C«HaO) - C = NOH precipitate with nickel salts in
I slightly ammoniacal solution. Its
(C,HaO) - C = NOH . advantages over dimethylglyoxime
(III) are its greater solubility in water,
its higher sensitivity (about 7
times greater) and the small nickel content of the complex
compouI\.d, thus giving .a larger weight of precipitate for a
.
given weight of nickel. It is, however, expensive:
a-Benzildioxime {C6HS'C( = NOH).C( = NOH).CoH 5 }, m.p.
237°C, is a more sensitive reagent for nickel than is dimethyl-
glyoxime, but has the disadvantage that it is less soluble than
the latter. It is therefore not usually employed except for
the determination of small quantities of nickel. Owing to
the voluminous character of the precipitate Ni(C l4l H ll Oi N2 )i
160 Quantitative Inorganic Analy~is
not more than about 0·02 g. of Ni can be accurately detet-

mined.
The reagent is prepared by dissolving 0·2 g. of the dioxime
in a mixture of 950 ml. of alcohol and 50 ml. of concentrated
q_mmonia solution.
For the determination of nickel a slight excess of a warm solution of the
reagent is added to the ammoniacal solution of the nickel salt and the whqle
heated on a water bath to coagulate the precipitate (ca. 2 minutes). The
precipitate is filtered on a weighed Gooch, sintered glass or porous porcelain
crucible, washed with 50 per cent alcohol, followed by hot water, dried to
constant weight at 1l0·C, and weighed as Ni(C 14H ll OzN t )2' ,
Nitrates and copper must be absent. The interference of ferrous iron and
chromium is prevented by the addition of sodium citrate or tartrate; of zinc
"and magnesium by ammonium chloride; of manganese by a slight excess of
acetic acid or citric acid and hydrazine (the latter to prevent oxidation of the
manganese); of cobalt by the addition of excess of the reagent.
B. Cupferron (the ammonium salt of nitrosophenylhydroxylamine).
This substance (IV) was intro-
/NO duced by O. Baudisch in 1909
O -N
",",ONH4
(IV)
for the separation of copper
from iron. It gives sparingly
soluble compounds with a number
of elements in which the ammon-
ium radical is replaced by One equivalent or-the metal in ques-
tion. Precipitation is made in the cold using a slight excess
pf a filtered 6 per cent aqueous solution of the re'agent. The
precipitate cannot be .satisfactorily dried, and is therefore
converted by ignition into the oxide of the precipitated
element and weighed as such. Precipitations are carried out
in strongly acid solutions (10 per cent by volume of sulphuric
acid or hydrochloric acid), and provide complete separations
of elements such as iron, vanadium, titanium, .tantalum- and
zirconiuII:l from aluminium, beryllium, chromium, manganese,
nickel, zinc and hexavalent uranium. Its principal applicc:-
tions in quantitative analysis are to the determination o!
iron, titanium, and zirconium.
A companion reagent to cupferron, a-nitroso-naphthyl-
hydroxylamine or neo-cupferron (IVA), was introduced by
O. Baudisch in 1937. Its proper-
ties are similar to those -of cup-
ferron, but the stability of its
aqueous solution is not quite so
high. Unlike cupferron, however,
it may be employed for the quan-
titative precipitation of small
(IVA) quantities of iron (and of copper)
in the presence of aluminium in strong mineral acid media;
it is thus useful for the direct determination of iron in
mineral waters and sea waters. *
C. 8-Hydroxyquinoline (oxine). This substance (V) ~as
introduced into quantitative analysis by R. Berg in 1927. It
CON OH
ON'\.~
0-0/ M""NU
/0-0
(V) (VI)
.forms sparingly soluble derivatives with metallic ions, which

have the composition M(CgHeON)2 (or VI) if the co-ordination
number of the metal is four (e.g., magnesium) and M(CgH60N) 3
if the co-ordination number is six (e.g., aluminium). There
are, however, some exceptions" e.g., TiO(C g H oON)2'
W0 2(Cg H 6 0N)2 and M002(CgHaON)2. By suitable adjust-
ment of the pH of the solution and by other methods (such as
the use of organic acids), numerous separations may be
carried out, for example, aluminium from beryllium in an
ammonium acetate-acetic buffer, magnesium from the
alkaline earth metals in ammoniacal solution, and the preci-
pitation of such metals as copper, cadmium, bismuth and zinc
under certain -experimental conditions.
The complexes are cOI)1pact and very easy to filter. They
may be weighed as such afMr drying at 100° to 105°C (hydrated
" oxinate") or at 130° to 140°C (anhydrous "oxinate "),
or they may be cautiously Ignited under a cover of oxalic acid
and weighed as oxide, or ~hey may be dissolved in hydro-
chloric acid and titrated with a standard solution of potas-
sium bromate (Section m, 138).
The reagent is prepared for use in either of the following
ways :-(i) Two grams of oxine are dissolved in 100 ml. 'of 2N
acetic acid, and ammonia solution is added drop wise until a
slight precipitate is just produced: the precipitate is dissolved
either by the addition of a little dilute acetic acid or by warm-
ing. This solution is stable more or l~ss indefinitely.
(ii) Two grams of oxine are.dissolved in 100 ml. of ethyl
* For further details, see the booklet Cupferran and Neo-cupferron (1938)
issued by the G. F. Smith Chemical Co., of 867 McKinley Avenue, Columbus,
Ohio, U.S.A. This finn supplies a particularly good quality of both cupferron
and neo-cupferron. The British,agents are F. W. Berk and Co. Ltd .• of 52-54
Leadenhall Street, London, E.C.3.
G
alcohol, methyl alcohol or acetone. This solution is stable
for about 10 days if not exposed to the light.
Notes on precipitations with oxine. Precipitation is effected under a variety
of conditions, e.g., in acetic acid or alkaline solutio.!t by an alcoholic 9' acetic
acid solution of oxine (see above). The reagent is added in excess as indicated
by the production of a yellow or orange-yellow colouration in acetic acicol or
alkaline media respectively. After heating fctr 2 to 3 minutes, the precipitate
is filtered through a filter paper, or a sintered glass (G3 or G4), porous porcelain
or Gooch crucible. The filtrate must be yellow or orange-yellow in colour,
thus proving the presence of excess of oxine. Frequently the precipitate is
washed with hot water until the filtrate is colourless and then with cold water.
When too large an excess of oxine is employed for precipitation. the filtrate
may become cloudy. This is usually due to oxine which is very much less
soluble in cold than in hot water, as is shown by heating the filtrate to the
boiling point when the solution becomes clear; a metallic" oxinate " would
not pass into solution on boiling.
Tj1e excess of the reagent from the frlttate and washings may be removed
by ether extraction in alkaline solution (PH 8 to 8·5) or, better, by evaporation
of the ammoniacal solution with repeated addition of ammonia solution (the
oxine is steam-volatile in a slightly alkaline solution). Determinations of
other elements by other methods can then be carried out and the concentrated
acid or ignition treatment, which is usual with many organic reagents (compare
dimethylglyoxime, Section IV, 89, g and f). is avoided.
D. a-Benzofnoxime (cupron). This reagent (VII) is principally
C6R5 - CR - OR C6R5 - CR - ~
(VII) I I /Cu (VIII)
CGR5 - C = NOR CGH~ - C = NO
applied in the determination of copper, with which it yields
a green precipitate (VIII) in faintly ammoniacil solution;
tile precipitate is dried at 105-115°C. Ions which are pre-
cipitated by ammonia solution are kept in solution by the
addition of sodium potassium tar.trate. Another important
application is to the determination of sexiyalent molybdenum
which is carried out in fairly strong acid solution; the
precipitated compound Mo 2 (C 14H u 0 2N)3 is best ignited at
500-525°C and weigJ:led as Mo0 3 •
The reagent is employed as a 2 per cent solution in rectified
spirit. .
E. Salicylaldoxime. This substance (IX) is a specific
reagent for copper if precipitation
O
~H • NOH is carried out in the presence of
acetic acid (compare, however,
Section IV, 32B). The greenish yel-
"OR Low precipitate Cu(C 7R 60 2Nh is
(IX) weighed after drying at 100-
105°C.
The reagerrt.is prepared by dissolving 1 g. of salicylaldoxime
in 4-5 ml. of cold 95 per cent alcohol and pouring the solution
into 95 m!. of water at a temperature not exceeding 80°C;
the mixture is shaken until clear and filtered, if necessary.
Another, and less expensive. method is to add 2·22 g. of
pure salicylaldehyde di5Solved in 8 ro!. of 95 per cent alcohol
to 1·27 g. of A.R. hydroxylamine hydrochloride dissolved in 2
. mI. of water: The resultant solution is diluted with 15 m!. of
95 per cent alcohol and stirred into 225 mI. of water at 80°C.
When cold, the solution is filtered, if necessary. This is an
approximately I per cent solution of salicylaldoxime.
Salicylaldoxime decomposes'slowly in solution; the reagent
cannot be kept for more than about 3 days.
F. a-Nitroso-,B-naphthol. This coin pound (Xl reacts with a
number of elements to form in-
soluble compounds in which the
}lydroxyl hydrogen is replaced by
an equivalent of the metal in
question. It was first recom-
mended for the separation of
cobalt from nickel by Ilinski and
V. Knorre in 1885. Precipitations (X)
are usually effected by adding the
cold reagent to a very dilute hydrochloric acid solution of the
metal, and heating to 60-70°C. The following elements are
quantitatively precipitated: iron, cobalt, palladium, silver
and gold; tin, titanium, vanadium, chromium, copper,
molybdenum and tungsten are partially precipitated. Separa-
tion is thus effected from aluminium, manganese, nickel,
_ zinc,. magnesium, and the alkaline earth metals. The
reagent is therefore far from being specific, and preliminary
separations are necessary'. Its principal use is for the
determination of cobalt in the presence of nickel; the red
precipitate, CO(C1oH602N) 3, obtained by precipitation of
cobaltous solutions, is contaminated with a little of a cobaltous
salts ~nd some of the reagent, so that it cannot be weighed as
such; moreover, ignition in air does not give a perfectly
pure cobalt oxide, C0 30 4 • However, if the cobaltous salt is
first'converted into cobaltic hydroxide Co(OH)a by means of
1OO-volume hydrogen peroxide and sodium hydroxide solution,
the precipitate dissolved in acetic acia and then precipitated
with the. reagent, then pure Co(C 1o H.0 2 N)s,2H 2 0 is pre-
cipitated, which may be weighed after drying at 130°C.
Palladium may be similarly precipitated and weighed as
Pd(C 1oH 6 0 2 N)2' after drying at 135°C; this method may lJe
used for separation from platinum.
The reagent is prepared by dissolving 4 grams of a.-nitroso-
,B-naphthol in 100 ml. of glacial acetic acid and then adding
100 ml. of hot distilled water'. The nltered solution must be
used immediately. The solution is C:.omparatively unstable
and should be prepared as required.
G. a.-Nitro-,B-Naphthol. This compound (XI) acts simi-
larly as a precipitant to a-nitroso-
;8-naphthol (e.g., it can be used for
NO 2 the determination of cobalt in the
O~
OJ (XI)
.
presence of nickel, zinc, mangan-
~se, aluminium, chromium, and
iron, and also for the determina~
tion of palladium in the presence
of platinum), but it has the great
advantage that its aqueous solu~
tion is stable. It should therefore be employed in preference
to the nitroso compound. Cobalt IS precipitated (hydrogen
peroxide and sodium hydroxide method) as Co(ClDHeOaN) a
and may be weighed as such after drying at 130°C. Palladium
is precipitated as Pd(C 1oH 60 aN)2 and may be weighed in
this form after drying at 130°-135°C.
Preparation of a-nitro-,S-naphthol from a-nitroso-,8~naphthol. Stir
a-nitroso-,8-naphthol with 10 times 'its weight of water and add slowly an
equal volume of nitric acid (tl 1·25). After a short time, the yellow needles of
the nitroso compound conglomerate into flocks and the odour of nitrous acid
becomes apparent. The reaction is complete after one hour. Wash the
yellowish-grey precipitate thoroughly with water. then heat it with 30 parts
of 1 per cent sodium hydroxide solution, and 1ilter. Acidify the clear solution
with acetic acid and filter of! the yellow precipitate of a-nitro-Jl-naphthol.
W ash ~ith water an'd recrystallise from hot alcohol. TJ:e pure compound has
m.p.l03°C.
The reagent is prepared by dissolving 2 g. of a-nitro-~~
naphthol in 100 ml. of cold glacial acetic acid, diluting with
100 ml. of hot water, and filtering.
H. Phenylthiohydantoic acid
(C6H5 - NH - C( = NH) - S - CH 2 - COOH). Certain
elements in ammoniacal or preferably in ammonium citrate
containing solution react with phenylthiohydantoic acid; an
equivalent of the element replaces the carboxyl hydrogen_.
Separation of cobalt and of copper and other elements of
Group II frpm aluminiqm, chromium, titanium, vanadium,
uranium, tungsten, molybdenum, zinc, manganese, calcium
and magnesium can thus be effected in one operation. Nickel
is partially precipitated, but the product is soluble in concen~
trated ammonia solution. The precipitate with cobalt is
The Theoretical Basis of Quantitative Analyst's 165
not of definite composition; it may be ignited to C0 3 0 4 •
The latter, however, is not quite pure; it may be fused with
potassium pyrosulphate, the cool melt dissolyed if! water, and
cobalt estimated by some other method (see Section IV, 33).
The reagent is employed' in the form of a 3 per cent aqueous
or alcoholic solution .•
I. Nitron (1:4 - diphenyl-3 : 5-endanilo - 4 : 5-dihydro-1 : 2 : 4-
triazole). Nitron (XII) is a strong
base; it precipitates nitrates as the
sparingly soluble crystalline nitrate
C2oH16N4,HN03 from solutions
slightly acified with sulphurIc acid.
The solubility of nitron nitrate
in very dilute sulphuric aci'd at
20°C is 10 mg. of salt per 100
ml. of saturated solution, which is
equivalent to 1·7 mg. of HN0 3.
The following acids form slighily
soluble salts with nit ron and therefore interfere with the
determination: perchloric, thiocyanic, hydriodic, hydro-
bromic, hydrochloric, chromic, chloric, and nitrous acids.
The reagent is employed as a 10 per cent solution in 5 per
cent acetic acid.
1. Tannin. Common tannin or tannic acid is essentially a
colloidal suspension of negatively charged particles capable of
flocculating the positively charged particles of certain inor-
ganic compounds, such as the sols of metallic hydroxides.
The separation of various elements depends to a large extent
upon the proper adjustment of the pH of the solution.
This reagent in the form of a freshly-.prepared.3 per cent or
10 per cent aqueous solution is useful for the separation of
some of the so-called rarer elements. It may be employed
inter alia for the quantitative determination of titanium and
tungstates, for the separation of aluminium, chromium, iron,
'etc., from beryllium, and of niobium from tantalum. In
most cases the element is ultimately ignited and weighed as
the oxide.
K. Phenylafsonic acid. (C oH 5 .AsO(OH)2)' This reagent

is particularly useful for the determination of tin; zirconium
and thorium are the only elements common~y associated with
tin which interfere. The-complex is ignited to and weighed
as stannic oxide.
166 Quantitativf- Inorganic Analysis
The reagent consists of a saturated aqueous solution of
phenylarsonic acid (an approximately 10 per cent solution).
L. Pyridine. Pyridine forms insoluble complexes with the
thiocyanates of cadmium, copper, nickel, cobalt, zinc and.man-
ganese; these have the general formula Me(CNS):,(C 5 H 5N) ..
(n = 4 for Co, Ni and Mn.; n = 2 fpr Cu, Cd and Zn)
(G. Spacu and co-workers, 1927-+). In practice, alkali
thiocyanate and a few ml. of pure pyridine are added
to the neutral or very faintly acid solution of tlie metal
ions. The complexes are readily filtered. They are washed
first with water, then with dilute alcohol (both contain-
ing a little alkali thiocyanate and pyridine), followed
successively by absolute alcohol and absolute ether, each'
containing a little pyridine. the precipitates are weighed
after drying in a vacuum desiccator for 5 to 30 minutes at the
laboratory temperature. The method is rapid, but, as is
evident, many ions interfere. The results for manganese are
not satisfactory because of the slight solubility of the complex
in the wash solutions. The method is applicable 'in the
presence of alkali. and alkaline earth metals and- of magne-
sium; considerable quantities of ammonium salts must be
absent.
Me Anthranilic acid. This substance (XIII) is employed in
the form of a 3 per cent soiution of
its sodium salt. It gives precipi-
/NH2 tates with neutral or weakly
O "'CO OR
(XIII)
acetic acid solutions of. zinc, co-
balt. nickel, cadmium, copper,
lead and mercuric mercury;
these have the general formula
• Me(C7H602N)2 (H. Funk and co-
workers, 1933-3"4). The precipitates are crystalline, insoluble
in water and alcohol, and non-hygroscopic;· they may
be dried at 105°-110°C. The reagent has been employed
for the deterinination of zinc, nickel, cobalt, cadmium..
and copper, but offers no great advantages over other
reagents_for these ions. Its use will therefore not be described
in ,Chapter IV. It may, however, be pointed out that it
is essential to employ pure anthranilic acid, m.p. 145°C, in
the preparation of the reagent.
N. Quinaldinic acid. This substance (XIV) gives insoluble
complexes with the ions of copper, lead, silver, cadmium,.
zinc, manganese, nickel, cobalt, and ferrous iron, and insoluble
The Theoreticai Basis of Quantitative Analysis' 167
basic salts with ferric iron, uranium (uranyl), aluminium,
chromium, beryllium, and titanium, In s1..l1uhuric acid solution,
however, and in the absence of
iron and zinc, quinaldinic acid is a
specific reagent for copper; the
. .
precIpItate h a.s the o f r m u 1a
CO
·--GeOH
Cu{ClOH s 02N)2' H 20 and may be N
weighed in this form after drying
at 125°C. The reagent is useful (f{IV)
for the rapid estimation of zinc
(as Zn(CloHs02N)2,H20) and of cadm_ium (as Cd(C 1o H s0 2N)2)
ip. the absence of interfering elements: in both cases the
precipitate is dried at 125°C.
The reagent is prepared by dissolving 2·50 g. of quinaldinic
acid or 2·82' g. of the sodium-s(j.lt in 75 ml. of water.
Note on the recovery of quinaldinic acid. The reagent is expensive, but may
be recovered after use. The precipitate and filtrate are best treated separately.
The filtrate from a determination is acidified with dilute su,lphuric acid and
precipitated as copper quinaldinate with copper sulphate solution; the
precipitate is filtered, washed, suspended in water, decomposed with hydrogen
sulphide, filtered from cupric sulphide, the filtrate evaporated to dryness, and-
the residue recrystallised from glacial acetic acid. The pure acid has m.p.
155°-156°C. The copper and cadmium complexes are decomposed with
hydrogen sulphide and treated as above. Zinc quinaldinate is best dissolved
in dilute hydrochloric acid or dilute sulphuric acid, the solution diluted until a
preclpitate just appears, and then precipitated with copper sulphate solution,
etc.
O. Pyrogallol. This compound (XV) yield insolubles
complex salts with antimony and
bismuth, and may be employed
for the quantitative determination
of these elements either alone or-
OH
/OH ,
in the presence of arsenic, lead,
etc.
The reagent" consists of a sQlu-
tion in air-free water. The pyro-
"OH
(XV)
0
gallol should have the correct·
m.p. (133°C), and should leav.e no residue when heated upon
platinum or silica.
P. Ethylenediamine (NH 2 .CH 2 .CH 2 .NH 2 ). Ethylenediamine

yields a complex cation with cupric ions:
Cu++ + 2NH 2 .CH2 .CH 2 :NH2
= [Cu(NH 2 ·CH 2·CH:.NH:)2J++ == [Cu en:J++ .
This reacts with the complex ions [HgI4r - or [CdI~1 - - to,
;
yield the insoluble complex salts [Cu en 2J [HgI,J and

[Cu enJ [CdIJ respectively:
HgCl 2 + 4KI = K 2 [HgI 4J + 2KCI;
K 2 [HgI41 + [Cu en J (NOah =
2 [Cu en 2] EHg!,] + 2KN0 3•
The complex salts are insoluble in water, 95 per cent alcohol

and ether, and hence' may be employed in the rapid deter-
mination of mercury, cadmium and copper respectively
(G. Spacu and co-workers, 1929 -+). The mercury complex is
stable in air and in a va_cuum, and its precipitation is unaffec-
ted by the presence of ammonium salts; a valuable rapid
method is thus available for the determination of mercury.
The cadmium complex has similar properties, but is slightly
soluble in the presence of an:tmonium salts or in strongly
ammoniacal solution; the experimental details for the
determination of cadmium are siryilar to those given for
mercury (Se~tion IV, 160) except that precipitation is effected
in neutral solution with reagent (a) (see below).
The process also permits of a quantitative separation of
copper and mercury. The acid solution of both metals is
treated with ammonia solution until nearly neutral and is
then rendered alkaline with ethylenediamine. The solutio_n is
heated to boiling, treated with excess of potassiUlb. iodide and
copper ethylenediamine nitrate solution and allowed to cool;
the mercury complex is filtered off and weighed (Section
IV,160). The copper is determined in another portion of the
solution as in Section IV, 19F.
The reagent may be prepared by either of the following
methods: (a) Heat an aqueous solution containing 1 part of
cupric nitrate and 2 parts of ethylenediamine on.a water bath
until a crust forms on the surface of the violet-blue solu-
tion. Allow to cool, filter off the separated crystals or
[Cu en2](NO ah,2H 2 0 at the pump, and wash them several
times with alcohol, followed by ether. A concentrated solu-
tion of this salt is used for precipitations. ,
(b) Treat a solution of c~pric sulphate with an aqueous
solution of ethylenediamine (5 to 6 times the theoretical
quarHity) until the dark blue-violet colouration, due to the
[Cu en 2J++ ion, appears and does not increase in intensity
upon further addition of ethylenediamine. The presence of
excess of the latter in the reagent has no harmful influence.
Here the reagent consists_of a solution orrCu en2]S04' and is as
satisfactory as (a(for determinations of mercury.
Q. p-Hydroxy-phenyl\\fSOnic acid. This valuable reagent (XVI),
introduced by C" T Simpson and
G. C. Chandlee in 1938, provides a
quantitative separation of titan- AsO(OH)2
o I
ium or zirconium. in acid solution
from ferrous and ferric iron,
chromium, aluminium, beryllium,
manganese, zinc, cobalt, nickel,
thorium, calcium, thallous thal- I
lium, cerous cerium, sodium, po- OR
tassium and ammonium as well as
from phosphate, chromate, per- (XVI)
manganate, molybdate, vanadate,
uranyl uranium and vanadyl vanadium. The final solution
should be about O'6N with respect to hydrochloric acid or
l'8N with respect to sulphuric acid. ,In mineral acid s'olution
of concentration not greater than O·5N, tin is quantitatively
precipitated. Thus tin, and also ceric cerium, interfere;
hydrogen peroxide must be absent. The titanium or zircon-
ium complexes are converted by ignition to the oxides, and
weighed 'as such. When both titanium and zirconium are
present, the Ti + Zr are first determined as the mixed oxides;
in a second determination use is made of the fact that the
zirconium alone is precipitated in the' pr~sence of excess of
hydrogen peroxide.
The reagent is'employed in the form of a 4 per cent aqueous
solution.
I, 63. VOLAmISATION OR EVOLUTION METHODS
Evolution or volatilisation methods depend essentially
upon the removal of volatile constituents. This may be
effected in several ways: (i) by simple ignition in air or in a
current of an indifferent gas; (ii) by treatment with some
chemical reagent whereby the desired constituent is rendered
volatile; and (iii) by treatment with a chemical reagent
whereby the desired constituent is rendered non-volatile.
The volatilised substance may be absorbed in a weighed
quantity of a suitable medium when the estimation is a direct
one, or the weight of the residue remaining after the volatili-
sation of a component is determined, and the proportion of the
constituent calculated from the loss in weight; the latter is
the indirect method. Examples of each of these procedures
are given in the following paragraphs; full experimental
details will.be found in Chapter IV.
,c*
] 70 Quantitative I nOl'ganic Analysis
The, determination of superficially bound moisture or of
water of crystallisation in hydrated compounds is usually
<:arried out simply by heating the substance to a suitable
temperature' and weighing the residue. I t is assumed of
course that the substance does not decompose upon heating
The water may also be absorbed in a weighed quantity of an
appropriate drying agent, such ·as calcium chloride, magne-
sium perchlorate or concentrated sulphuric acid.
The determination of carbon dioxide in carbonate-contain-
ing materials may be effected by treating the sample with
excess of acid and absorbing the carbon dioxide in alkaline
absorbent, .such as soda lime, soda lime-asbestos, or sodium
hydroxide-asbestos (" ascarite "). The gas is completely
expelled by heating the solution and by passing a current of
purifie9- air through the apparatus; it is, of course, led
through a drying agent to remove water vapour before passing
to the carbon dioxide absorption apparatus. The gain in
weight of the latter is due to, carbon dioxide.
In the determination of carbon in steels and alloys, the
substance is burnt in pure oxygen in the presence of catalysts
and the carbon'dioxide absorbed as in the previous example.
Precautions are taken to, remove other vOlatile constituents,
such as sulphur dioxide. This method is emp~oyed in the
determination of carbon and hydrogen in organic compounds;
the sample is burnt in a controlled.stream of oxygeil, and the
water and -carbon dioxide are absorbed separately in an
appropriate absorbent, e.g., in calcium chloride saturated with
~arbon dioxide and in soda lime (or" sofnolite "-see Section
IV,76). •
A similar ,determination is the. estimation of ammonia.
By heating an ammonium compound with excess of caustic
alkali, ammonia is quantitatively evolved and is absorbed in a
measured excess of -a standard solution of an acid. The
excess of the latter is then estimated by titration with
standard alkali.
'(NH')2S04 + 2NaOH = Na 2S0 4 + 2H 2 0 + 2NH s.
This method may be adapted to the determination of nitrogen
in many, but not all, complex organic compounds .and to
substances containing nitrogen in the organic form. The
latter are heated with excess of concentrated sulphuric acid,
preferably in the presence of a catalyst such as selenium
oxychloride or copper sulphate, whereby the organic nitrogen
is converted into ammonium bisulphate. Some potassium
sulphate is usually added to raise the boiling point of the
acid and thus accelerate the conversion.. The acid is then
neutralized, and the ammonia liberated wit)J. alkali as above.
This is the Kjeldahl method for the determination ot nitrogen.
Some elemants, such as- sodium and potassium, may be
determined in solution or when combined by evaporating to
dryness with sulphuric acid; the residual sulphate is then
weighed.
Interfering metals m)lst, of course, be removed first.

An example of a related kind is the determination of pure
silica in an impure ignited silica residue. The latter is treated
in a platinum crucible with a mixture of sulphuric and hydro-
fluoric e.cids;- the silica is converted into the volatile silicon
tetrafluoride.
SiO z + 4H,F ~ SiF4 + 2H zO.
The residue consists of the impurities, and the loss in weight
of the crucible gives the amount of pure silica present. .
ELECTRO-ANALYTICA,L MEmODS
I. 64. Theory of electro-analysis.-In electro-analysis the
element to be determined is deposited electrolytically upon a
suitable electrode. Filtration is thus avoided and coprecipi-
tation, if the experimentaf conditions are carefully controlled,
is very rare. The method, when applicable, has ma_ny advan-
tages, and we shall therefore study the theory of the' process
in order to understand how and when it may be applied.
Electro-deposition is controlled by Ohm's law and by
Faraday's two laws of electrolysis (1833-34). The Jatter
state:
(1) The amounts of substances liberated at the electrodes are
proportional to the quantity of electricity which passes through
the solution.
(2) The amounts of different substances which are deposited,
or liberated, by the same quantity of electricity are proportional
to their chemical equivalents.
It follows from the second law that when a given current is
passed in series through solutions containing, say, copper
sulphate and silver nitrate respectively, then the weights of
copper and silver deposited will be in the ratio of their
equivalents, viz., 107·88: 63'57/2.
Ohm's law expressed the relation between the three funda-
mental quantities, current, electromotive force and resistance.
172 Quantitative I-norganic Analysis
The current C is directly proportional to the electromotive Jorce
E and inversely proportional to the resistance .{?,
i.e., C = E/R.
Electrical units. The unit of current is called the ampere.
and is defined as that current which flowing for one ·second
will cause the deposition of 1'~1800 miliigrams of silver, or~
0·3294 milligram of copper, from solutions of their respective
salts.
The unit of electrical resistance is the ohm. The Inter-
national ohm is the resistance at O°C of a column of mercury,
of uniforp:l: cross section. IOt}'3 cm. long, and weighing 14·4521
grams.
The unit of electromotive force (e.m.f.) is termed the volt,
and is the difference of electrical potential required to main-
tain a current of one ampere thr0ugh a system having a
resistance of one ohm.
The unit quantity of electricity is the coulomb, and is
defined as the quantity of electricity passing when one ampere
flows for one second. Each coulomb will deposit 1·11800 mg.
of mercury, or 0'3294 mg. of copper.
The weight of an element liberated by the passage of one
coulomb of electricity (or one ampere for one second) is called
the electrochemical equivalent of the element. Tpe equiva-
.lent weight of silver is 107'88, hence 107'88/0'0011180, i.e.,
96,494 coulombs will be required to liberate one gram equiva-
lent of Silver. The value generally employed is 96,500
coulombs. and this is termed a faraday (F). The second law
of electrolysis may be stated in the form that one faraday of
electricity will liberate one gram equivalent of any substance
at each electrode.
I, 65. Decomposition potential.-If a small voltage of,
say, 0·5 volt is applied to two smooth platinum electrodes
immersed in a solution of molar sulp1Wric acid, then an am-
meter placed in the circuit will at first show that an appre-
ciable current is flowing. Its ,strength, as shown by the
ammeter, decreases rapidly and after a short time becomes
virtually equal to zero. If the- applied voltage is gradually
increased, there is a slight increase in the current until, when
the applied voltage reaches a certain value, the current sud-
denly increases rapidly with increase in the e.mJ. It will be
observed, in general, that at the point at which there is a sudden
iilcrease in current bubbles of gas commence to be freely
evolved "at the electrodes. The experiment may be carried
out by means of the apparatus shown diagrammatically in
Fig. 1-28. A storage battery C is connected across a uniform
resistance wire AB, along c
which a contact maker D 11111-----""
can be moved; the fall
'of potential between A
and D can thus be varied AI-'+~_ _ _ _ _ _ _ _7D_ __'B
gradually. Two smooth
platinum electrodes are
immersed in molar sul-
phuric acid in the cell E.
V is a suitable voltmeter
placed between the two
electrodes across the cell; E
M is a milli-ammeter and FIG. 1-28.
F is a switch. When the
sliding contact is near to A, only a small potential is applied
to the electrodes of the cell; the fall of potential across the
cell and the current flowing through it are read off on the
. instruments Van d A
respectively. The applied
. voltage is slowly increased
by moving D towards B,.
and the readings in the
voltmeter and ammeter
noted after allowing a
short time for the values
to become steady. Upon
DecompDsltion
: / Vollage
plotting the cur r e n t
against the applied volt-
Voltage across th'e Cell . age, a curve similar to
FIG. 1-29.
that shown in Fig. 1-29 is
obtained; the point D at which the current suddenly increases
is clearly shown, and in the instance under consideration is
about 1·7 volts. The voltage at the point D is termed the
decomposition potential and it is at this point that the evolu-
tion of both hydrogen and oxygen in the form of bubbles is
first observed. We may define the decomposition potential of
an electrolyte as the minimum external voltage that must be
applied in order to bring about continuous electrolysis. *
• Strictly speaking, the voltage across the cell, measured by the voltmeter
V, is greater. than the aecomposition potential by the quantity cr, where c
is the current flowing in amperes and r is the resistance in ohms. However,
since c is very small and r is not very large, the quantity C1' may be neglected,
and the voltage measured on V taken as the decomposition potential.
17-1- Quantita! ive Inorganic Analysis
If the current is broken after the e.m.f. has been applied, it
will be observed that the reading on the voltmeter V is at
first fairly steady and then decreases, more or less rapidly,
tq zero. The cell E is clearly behaving as a source of current
awl is said to exert a back or counter or polarisation e.m.f., since
the latter acts in a direction opposite to that of the applied e.m.f.
This back e.m.f. arises from the accumulation of oxygen
and hydrogen at the anode and cathode respectively; two
gas electrodes arc consequently formed and the potential
difference between them opposes the applied e.m.f. When
the primary current from the battery is shut off, the cell
produces a moderately steady current until the gases at the
electrodes are either used up or have diffused away; the
voltage then falls to zero. This back e.m.f. is present even
when the current from the battery passes through the cell
and accounts for the shape of the curve in Fig. 1-29. It is
evident that the minimum value of the counter e.mJ. may be
computed for it is equal to the algebraic difference of the
electrode potentials which exist at the anode and cathode
respectively. This calculation will be referred to again in
the succeeding paragraphs.
Some results of decomposition potential measurements,
made by Lc Blanc in 1891-93, are collected in Table XXI.
TABLE XXI. DECOMPOSITION V.OLTAGES OF N-SOLlJTIONS BETWEEN
SMOOTH PLATINUM ELECTRODES
Substance I ~com!_~ ~o!!.age :1__ -=Ubstanc_8__ I Decomp. Voltage

ZnSO, 2·35 NiCl, 1·85
NiSO. 2·09 ZnBr2 1·80
CdSO. 2'03 CoCl, 1·78
Cd(NO a), 1-1)8 Pb(NOal. 1·52
CoSO. 1-1)2 AgNO, 0·70
CdCl. 1·88
HNO a Hj9 HBr 0'U4

H 2 SO, Hi7 HI 0·52
HaPO, 1·70 NaOH 1·69
HClO. I·H5 KOH 1·67
HCI j·:H NU.oH 1·74
It will be observed that whereas the decomposition potentials

of salt solutions vary considerably, those for the acids and
alkalis, with the exception of the halogen acids, are all
approximately 1'7 volts. It was therefore concluded that
the same electrolytic process occurs with these acids and
The Theoretical Basis of Quantitative Analysis 17 fi
bases; this can only be the discharge of hydrogen and
hydroxyl ions at the cathode and anode respectively leading
to the liberation of l}ydrogen and oxygen.
2H-r t~ 2€ ~ H 2•
2€ +- 20H -~ H 2 0 + 1/20 2 ,
~
With the halogen acids in' N solution, the halogen and not
oxygen is liberaterl at the anode since the discharge of the
halogen can occur more readily than the hydroxyl ions.
I, 66. Electrode reactions. ~In electro-analysis we are
largely concerned with the electrolysis of salt solutions, and it
is therefore proposed to study the reactions which take place
at the electrodes in some detail. Let us consider the electro-
lysis of a molar solution of zinc bromide between smooth
platinum electrodes. The application of a voltage will result
in'the deposition of zinc on the cathode (thus producing a zinc
electrode) and bromine at the anode (thus producing a bro-
mine electrode). The reaction at the cathode is :
Zn++ I- 2€ ~ Zn,
i.e., an oxidation (Section,I, 23), and that at the anode is:
2Rr- ~ Br2 + 2€,
i.e., a reduction. Thus oxidation occurs at the cathode and
reduction at the anode. \Ve may calculate the potential
at the cathode at 25°C from the formula (Section I, 43) :
E,uthode = EOZll + o.~~! log [Zn++] =-= EO zn '
since [Zn H] -= 1. At the ~tnode :
0·0591
E anod , ~ E"'rlr2 - 1 log [Br ] = EO nr.,
since [Br-] = 1. The e.m.f. of the resultant cell will therefore
be:
EOzll - EO Br • = 0·76 ~ ( - 1'07) = 1·83 volts.
This value is in agreement with the observed decomposition
potential (see Table XXI).
For a similar electrolysis of normal zinc sulphate solution,
the reaction at the anode is :
20H' _,_ H 20 + 1/202 j- 2 €,
for the hydroxyl ion is more easily discharged than the sul-
phate ion. An oxygen electrode is thus produced at the anode.
The potential of a reversible oxygen electrode is a function of
1.'16 Quantitative Inorganic Analysis
the PH of the solution, and has heen shown to be given by t,he
expression:
E = EOo> + 0'0591 log [H+].
E002 is 1·23 volts, and the oxygen potential o('a solution
normal with respect to hydrogen ions is accordingly 1·23
volts.
In general, it may be stated that the theoretic:!! back or
polarisation e.m.f. Eb is given by :
Eb = Etclhode - E allodl ,
where Ecathode and Ea>lod, are calculated as already des-

cribed (Section I, 43).
I. 67. Overvoltage.-It has been found by experiment that

the decomposition voltage of an electr9lyte varies with the
nature of the electrodes employed for the electrolysis and is,
in many instances, higher than that computed from the differ-
ence of the reversible electrode potentials. The excess voltage
over the calculated back e.m.f. is termed the overvoltare.
The overvoltage may occur at the anode as well as at the
cathode. The decomposition volt<;tge ED is therefore:
ED = Ecalhode + E o.c. - (Eanode + E o.a.) ,
",here E n.c• and Eo .•. are the overvoltages at the cathode and
anode respectively.
The overvoltage at the anode or cathode is a function 6f the
following variables:
1. The nature and the physical state of the metal employed
for the electrodes.
2. The physical state of the substq_nce deposited. If it is
a metlil, the overvoltage is usually small; iht is a gas, such
as oxygen or hydrogen, the overvoltage is relatively great.
3. The currerit den.sity employed. * For current densities
up to 0·01 amp.jcm.2, the increase in overvoltage is very
rapid; above this figure the increase in overvoltage continues,
but less rapidly. ~
4. The change. in concentration, or' the concentration
gradient, existing in the immediate vicinity of the electrodes;
as this increases, the overvoltage rises. The concentration
gradient depends upon the current. density, the temperature
and the rate of stirring of the solution.
,.. The current density; C.D., is defined as the current per unit a'rea of elec-
trode surface. It is. generally expressed in amperes per square centimeter
NDl' or per square deeiIIteter (100 sq. ems.). ND,oQ'
The Theoretical Basis of Q1tantt"tative Aualysis 177
As a rule, metals exhibit small overvoltage effects. Oxygen
shows an overvoltage of about 0·45 volt at the anode in acid
solution, and is of the order of 1 volt in alkaline solution on a
smooth platinum surface with current densities of 0·02 to
01()3 amp./cm. 2 • The hydrogen overvoltage is of great
importance in. electrolytic determinations and separations.
The values in molar sulphuric acid are shown diagrammatic-
ally in Fig. 1-30. In alkaline solutions the hydrogen over-
voltage is slightly higher (0·05 to 0·3 volt) than in acid
solutions.
1'4
Pb
eu
1'2
II V
1·0 /
yl-'"
v I- - Ag
/ 7 Au
il ...- -
-- - -
~ 1--.:-:.;:1--
yt srntt~-
il / r
...-l -
I /
0'4
'/ V
V
/
j
0·2
l Pt black
...:
0·5 1·0 .1·5
Cllrrent density. amps. per sq. em.
FIG. 1-30.
I, 68. Completeness of deVOsition.-Let us consider the

variations in e.m.f. at the cathode during the deposition of a
metal in an electrolytic deter:mination. Let the ionic con-
centration at the commencement of the estimation be c;.
For a bivalent metal, e.g., coppe~, the cathode potential at
25°C will be :
0.0591
E~11 + ~ 1og c, - E~1l + 0·030 lo~:crvolts.
If the ionic concentration is reduced to one-ten thousandth of
its original value (i.e., to secure an accuracy of 0·01 per cent
in the det~rmination), the new cathode potential.will be:
E~II + 0·030 log c. X 10- 4
=E~II + 0'030 log Cj + 0·030 log l(~-4
= (E~u + 0·030 log c.) - 4 X 0·030
= Potential at commencement of deposition - 0·102 volt.
This is independent of the value of Ci, and hence whenever
the ionic concentration is reduced to one ten thousandth of its
initial value (this may be regarded as the ultimate limit of the
ele~tro-analytical estimation, although for most purposes
an accuracy of 0·1 per cent is regarded as sufficient), the
potential is altered by 4 X 0·0591/2 = 0'1~0 volt for a
bivalent ion. For a univalent ion, the change is 4 X 0'0591/1
= 0'236 volt, and for a terval~nt ion it is 4 X 0'0591/3 =
0·079 volt. Since the back e.m.f. is produced by, a cathode
metal acting as the negative pole of a cell, the positive pole
being, say, oxygen, it follows that the back e.mJ. will become
greater as the metal becomes more negative during the course
of the analysis. Otherwise expressed, _the decomposition
potential increases as the deposition of the metal proceeds.
For quantitative deposition the applied e.m.f. :.;nust equal or
exceed the decomposition voltage when the cohcentration of
the given cation is negligibly small (say 10-( of the initial
value).
It is important to know the conditions for the deposition
of the metal in preference to hydrogen in an electrolysis.
The condition for the deposition of the metal is evidently that
the 'potential difference betweeIJ. the electrolyte and the
cathode, E 50loll'on. metal, must be less than the reversible
deposition. of hydrogen plus the overvoltage of hyq.rogen
(OR,) for the metal under-consideration. The relationship
may be expressed in several ways:
E.olulion.-metal < (E R +. H2 + OR.),
< (-E H •• H + + OR.),
< .(0·059 pH + OR.) (at 25°C).
If the cathode potential, Emelal. ioft, is substituted for

Esolulwn. metal, the expression becomes:
E melal• ion < (0·059 pH + QH2)'
I, 69. Electrolytic separation of metals.-It will be clear
from what has been stated in the preceding Section that the
initial deposition potentials of two metals must differ by at
least 0·25· volt f0r a virtually quantitative separation to be
theoretically possible. This minimum value would require a
very precise control of the potential' drop at the cathode;
for most practical -purposes, the difference in potential should
be at least 0·4 volt. From an inspection of Table XXI it is
apparent that certain metals can be separated electrolytically
with great ease, for example, copper from zinc, nickel, cobalt
or lead, silver from copper, etc. When, however, the stan-
da~d potentials of the two metals differ only slightly, the
electro-separation is more difficult. The obvious me'thod
is to alter the electrode potential of the one of the metals in
some way. This is most simply achieved by decreasing the
ionic concentration of the ion being discharged by incorporat-
ing it in a complex ion of small instability constant (Section
I, 12). The decompositio1n potential of the metal forming a
complex ion is thus raised. Furthermore, the overvoltage at
the small ionic concentration is also usually increased. A
possible consequence of changing the electrode potential in
this way is that a metal which. in simple ionic solution is
liberated at a lower voltage than another metal, may exhibit
the reverse behaviour in a complex-forming medium..
Some results, largely due to Foerster (1906), for the deposi-
tion potentials (=' - Ecathode) of some metals in simple and
alkali cyanide solutions are given in Table XXII.
TABLE XXII. DEPOSITION POTENTIALS En OF SOME METALS IN SIMPLE AND

IN ALKALI CYANIDE SOLUTIONS.
ED for O·IM Concentration of Excess of KCN per 0·1 Mol

Metal solutions of of Simple Metallic Cyanide
• M++ ions (vo/t)
O'2M 0·4M II M
Zn + 0·79 + 1·03 + l-lS + 1·23
Cd + 0·44 + 0·71 + 0'S7 + 0·90
Cn ,_ 0·31 + 0·61 + 0·96 + 1-17
An interesting application of these results is to the direct

quantitative separation of copper and cadmium. The copper
is first precipitated in acid solution; the solution is then
neutralised with pure aqueous sodium hydroxide, potassium
cyanide is added until the initial precipitate just redissolves,
and the cadmium is deposited electrolytically. Another
application is the simultaneous of zinc and copper (brass)
in the presence of molar potassium cyanide.
I, 70. Character of the deposit.-The ideal deposit Jor
analytical purposes should be coherent, dense and smooth;
in this form it is readily washed and the solution rinsed off
without loss. Flaky, spongy, powdery or granular deposits
adhere only' loosely to the electrode, and for this and other
reasons should be avoided.
As a rule, more satisfactory deposits are obtained when the
metal is deposited from a solution in which it is present as
complex ions rather than as simple ions. Thus silver is
. obtain~d in a more adherent form from an argentocyanide
solution than. from silver nitrate solution. Nickel when
deposited from solutions containing the complex ion
[Ni(NHa)6J++ is in a very satisfactory state for drying and
weighing. Mechanical. stirring often improves the character
of the deposit since large changes of concentration at the
electrodes are reduced, i.e., concentration polarisation is
brought to a minimum.
Increased current density up to a certain critical value
leads to a diminution of grain size of the deposit. Beyond
this value, which depends inter alia upon the nature of the
electrolyte, the rate of stirring and the teplperature, the
deposits tenet to become unsatisfactory. At sufficiently high
values of the current density, evolution of hydrogen may
occur owing to the depletion of metal ions near the cathode:
If an appreciable evolution of hydrogen occurs, the deposit
will usually become broken up and irregular; spongy
and incoherent deposits are generally obtained under
such conditions. For this reason the addition of nitric acid
or ammonium nitrate is often recommended in the determina-
tion of certain metals, such as copper; bubble formation is
thus prevented. The nitrate ion reacts with the hydrogen
as it is formed at the cathode:
NO a- + 10H+ + 8€ = NH.+ + aHIIO,
or HNO s + 4Ha = NHs + 3H 20.
For analytical purposes the current densities usually range
between 0·1 and 0·5 ampere per square decimeter for electro-
lysis made in still electrolytes; somewhat higher values are,
however, occasionally employed.· .
Raising the temperature, say, to between 60°C and 90°C,
often improves the physical properties of the deposit. This is
due to several factors, which include the decrease in resistance
of the solution, increased rate of diffusion, and changed
ionisation and oyervoltage effects.
In practice, two methods of electrolysis are utilised. In
the first method stationary electrodes are used and the solu-
tion is not stirred; small current densities must of necessity
be applied in order to secure a coherent deposit, and the pro-
cedure is a slow one (slow electrolysis). In the second method,
which is largely superseding the first, the solution is rapidly
stirred (rapid electrolysis). Various devices are employed for
stirring. An independent mechanical stirrer may be used,
but it is more usual to have a rotating anode, which may
consist, after. Sand, of a platinum wire gauze cylinder sur-
rounded by a similar (but stationary) cylinder, which con-
stitutes the cathode, the intervening space being small (3 to
5 mm.). A very much higher current density may then be
applied (ND 100 ' i.e., a carre.nt density per 100 sq. cm., of 5
to 10 amps.) without seriously affecting the purity or. the
physical character of the deposit. The stirring results in a
liberal supply of metal ions always being present near the
cathode, and consequently the cprrent is principallY,used in
the deposition of the metal or other plate. A considerable
saving of time is thus effected, and this accounts for the
popularity of the method. It must be emphasised that when
the electrolysis is complete, the current must not be switched
off as long as the electrodes are in the solution. If the current
were broken, the counter e.m.f. would come into play and this
would cause part of the metallic deposit to pass back into
solution.
I. 71. SEPARATION BY MISCELLANEOUS PHYSICAL

METHODS
A. Extraction of solid mixhtres by solvents. This methdd
is based upon the dissolution of the desired constitnent in
some liquid, usually an organic solvent, in which the solubility
{)f any other substances present is so small as to be negligible.
Thus in one method for the separation of the alkaline earths,
the metals are converted into. the anhydrous nitrates, and
use is made of the fact that calcium nitrate is soluble whilst
strontium and barium nitrates are insoluble in absolute
alcohol. After extraction of the calcium nitrate with alcohol,
say, in a Sohxlet apparatus, the residue is converted into the
chlorides by repeated evaporation with hydrochloric acid.
182 Qutmtitative Inorganic Analysis
Strontium chloride may then be extracted w!th absolute
alcohol, and barium chloride remains behind.
Other examples include the separation of.lithium chloride
from sodium and potassium chlorides by extraction with
normal butyl alcohol, amyl alcohol, a mixture of etl'ler and
absolute ethyl alcohol, acetone, or dioxan (diethylene dioxide),
and the separation of sodium and potassium perchlorates by
extraction of the soluble sodium salt with a mixture of equal
volumes of ethyl acetate and normal butyl alcohol. Strontium
may be separated from calcium by making use of the fact
that calcium nitrate is soluble in anhydrous acetone whilst
strontium nitrate is insoluble.
B. Extraction of solutions by solvents. Partition or dis-
tribution method. A mixture of two substances in a given
solvent may often be separated virtually completely if a
solvent can be found, which is immiscible with the first, and
in which one of the substances is readily soluble and the other
practically insoluble. Thus if a mixture of ferric and alu-
miniqm chlorides in 6N hydrochloric acid be shaken with
ethyfether saturated with, hydrogen chloride, the ferric chloride
passes largely into the ether layer. By repeating the extrac-
tion with fresh portions of ether, the separatio~ may be made
practically complete. The results for a,number of chlorides
are collected in Table XXIII below_,
TABLE XXIII. EXTRACTION OF METAL CHLORIDES BY ETHYL ETHER FROM
6N HCI SOLUTIONS.
Element IPer cent extracted

at each extraction
Element Per cent extracted
at each extraction
Sb (SbCI.) 6 Fe (FeCI.) 99
Sb (SbCI.) 81 Hg (HgCI.) 0·2
As (AsCI.) '68 Mo (MoCl.) 80-90
Cu - 0·05 Te (TeCI.) 34
Ga 97 TI (TlCI.) 90-95
Ge 40-60 Sn ~SnCI.) 17
Au (AuCl.) 95 Sn SnCl.) 15-30
Ir (IrCI,) 5, Zn 0·2
I
.
AI, Bi, Be, Ca, Cd, Cr, Co, Fell, Pb, Mn, Ni, Os, Pd, rare earths,
Rh, Ag, Th, Ti, W, D, and Zr are not extracted.
-
The ethyl ether extraction method may be employed to
separate ferric iron from elements like aluminium, chromium,
The Theoretical Basis of Quantitative Analysis ' 183
cobalt, -nickel, vanadium, molybdenum and titanium, which
accompany it in many alloys. The procedure is to:evaporate
the hydrochloric acid s,olution of iron, etc., which should be
free from alkali chlorides (these are insoluble in ether-hydro-
chloric acid) and from substances that decompose ether
(chlorine, nitric acid, etc.), to tccke up the residue with hydro-
chloric acid (sp. gr. 1·1), and to transfer it with the aid of this
acid to a separating funnel. The cold acid solution is ex-
tracted three times by shaking with ethyl ether: between
each extraction the ether layer is allowed to separate before
the lower layer is drawn off for re-extraction. The iron is
removed from the ethereal extract by shaking it with water
and drawing off the lower aqueous layer. Heat is generated
by the mixing of the ether and ferric chloride-hydrochloric
acid solution so that cooling the mixture under the tap or
in ice is essential. The liberation of heat is considerably
reduced 'Qy the use of. concentrated hydrochloric acid saturated
with ether for dissolving the syrupy residue of ferric chloride,
etc. (100 ml. of hydrochloric acid, sp. gr. 1'19, dissolve 150 ml.
of ethyl ether; if the sp. gr. of the acid is 1·1 only 30 ml. of
ether is dissolved).
Improved results are obtained with the less-volatile isopro-
pyl ether (R. W. Dodson, F. J. Forney and E. H. Swift,
1936). In particular, the extraction of ferric chloride from
aqueous hydrochloric acid solutions is more efficient for a
wider range of acid concentration (6·5 to g'5N; the optimum
concentration is 7·75 to g'ON and the efficiency is 99·9 per
cent after 3 extractions) than for ethyl ether, and furthermore
the volume changes of the isopropyl ether and aqueous layers
are "Smaller and no reduction of ferric iron occurs when both
isopropyl alcohol and peroxide are present in modera'~e
amounts. Satisfactory separations of iron from copper,
cobalt, nickel, manganese, tinc, aluminium, chromium,
quadrhralent vanadium, titanium and sulphur (as sulphuric
acid) can be obtained. Large amounts of quinquevalent
vanadium are extracted; phosphoric acid and hexavalent
molybdenum pass into the ether-layer with ferric iron.
Gold chloride may also be extracted in a similar manner
using ethyl acetate as solvent. An interesting application
is to the determination of molybdenum in steel; the brown
colouration produced by the 'formation of molybdenum
thiocyana.te is extracted with cyclohexanol (see'Section V. 16).
The theory of the process is as follows. If a solute is shaken
with two immiscible solvents, and its molecular state i's the
1~4 Quant£iative Inorganic Analysis
same in both, then it has been found that at any given
temperature:
Concentration of solute in solvent A C K
- - - - - - - - - - - - = - A = a constant = I
Concentration of solute in solvent B CB
where K is the distribution or partition coefficient. Thus K
for iodine distributed between carbon bisulphide (A) and
water (B) is' about 410 at 18°C. The problem in practice is to
determine what is .the most efficient method for removing a
substance quantitatively from solution. It' can be shown'"
that if TV ml. of, say, an aqueous solution containing Xc
grams of a substance be extracted n times with L ml. portions
of a given solvent, then the weight of the substance Xn remain-
ing in the water is given by the expression:
X" = Xo (KWKW)R
+L
where K is the distribution coefficient between water and the
given solvent (compare Section I. 60). It follows, therefore,
that the best method of extraction with a given volume of
extracting liquid is to employ several fractions of the liquid
rather than to utilise the whole quantity in a single extraction.
Let us take a particular ~xample. Let us spppose that. we
shake 50 ml. of water containing 0·1 g. of io?ine with 25 mI.
of carbon tetrachloride. 'l:he distribution coefficient of
iodine between water and carbon tetrachloride at the ordinary
laboratory temperature is 1/85, i.e., at equilibrium the iodine
concentration in the aqueous layer is 1/85th of that in the
carbl)ll tetrachloride layer. We will compute the weight 'of
iodine remaining in the aqueous layer after one extraction
with 25 ml. and also after, say, three extractions with 8·33
ml. of the solvent by application of the above formula. The
former can also be simply computed as follows. If Xl g. of
iodine remains in the 50 mL of water, its concentration is
x l /50 g. per ml.; the concentration in the carbon tetra-
chloride layer will be (0·1 - Xl) /25 g. per ml. Hence:
x l /50 1
= -5' or Xl = 0,00230 g.
(0·1 - Xl ) / 25 8
The concentration in the aqueous layer after three extractions
with 8,33 ml. of carbon tetrachloride is given by :
(1/85) X 50
X3 = 0·1 ( (50/85) + 8.33
)3
= 0·0000145 g .
... For derivation, see Taylor., A Treatise on Physical Chemistry, 1930, I,
p. 487 (Macmillan and Co. Ltd.).
The extractioJil may therefore be regarded as virtually
complete. .
If the distribution coefficient is favourable, a separating
funnel may be employed; if it is comparatively small, some
continuous form of extraction apparatus is to be preferred.
I, 72. ERRORS IN QUANTrr,ATIVE ANALYSIS

When a .quantity is measured with the greatest" exactness
that the instiument, method and observer .are capable of,
it is found that the results of successive determinations differ
among themselves to a greater or lesser extent. The average
value is accepted as the most probable. This may not always
be the true value. In some cases the differe:w;e may be small,
in others it may be large; the reliability of the result depends
upon the magnitude of this difference. It is therefore of
interest to enquire briefly into the factors which affect and
control the trustworthiness of chemical analysis.
The absolute error of a determination is the difference
between the observed or measured value and the true or most
probable value of the quantity measured. The absolute error
is a measure of the accuracy of the measurement. The
accuracy of a determination may therefore be defined as the
concordance between it and the true or most probable value.
The relative error is the absolute error divided by the true or
most probable value; it is usually expressed iIi terms of per-
centage or in parts per thousand. The true or absolute value
of a quantity cannot be established experimentally, so that
the observed result must be compared with the most probable
value. With pure substances the quantity will ultimately
depend upon the "atomic weights of the constituent elements.
Determinations of the atomic weights have been made with
the utmost care, and the accuracy obtained far exceeds that
attained in ordinary quantitative analysis; the analyst
must accordingly accept their reliability. With natural or
industrial products, we mu,?t accept provisionally the results
obtained by analysts of repute using carefully tested methods.
If several analysts determine the same constituent in the same
sample by different methods, the most probable value, which
is usually the average, can be deduced from their results.
In· both cases the establishment of the most probable va:lue
involves the application of the theory of probability and the
concept of precision.
The agreement between a series of results is measured by
computing their'mean deviation. This is evaluated- by d~ter
mining the arithmetical mean of the results, then calculating
the deviation of each individual measurement from the mean,
and finally dividing the sum of the deyiations, regardless of
sign, by the, number of measurements. The relative. mean
deviati()n is the meap. deviation divided by the mean. This
may be expressed in terms of percentage or in parts per
thousand. An example will'm3;ke this clear.
Example. Th~ percentages of a constituent A in a compound AB
were found to be 48'32, 48·36, 48'23, 48·Il and 48·38 per cent.
Calculate the mean deviation and the relative mean deviation.
Results Deviations
48·32 0·04
48·36 0·08
48·23 0·05
48·11 0·17
48,38 '0·10
5)241'40 5)0'44
Mean = 48·28 Mean deviation '= 0·09
Relative mean deviation = 0·09 X 100/48·28 = 0·19·per cent

= 7'9 parts per thousand.
Precision may be defined as the concordance of a series of
measurements of the same quantity. The mean deviation
or the relative mean deviation is a measure of precision. In
quantitative analysis, the precision of measurements rarely
exceeds 1 to 2 parts per thousand.
Accuracy expresses the correctness of a measurement, and
precision the reproducibility of- a measurement. Precision
always accompanies accuracy, but a high degree of precision
does not imply accuracy. This may be illustrated by an
example.
Example. A substance was known to contain 49·06±0·02 per cent
of a given constituent A. The results obtained by two observers
using the same substance and the ,same general technique were:
Observer 1. 49'01; 49·21; 49'08. Mean = 49·l() per cent.
Relative mean error = (49·10 - 49·06) /49·06-- 0·08 per cent.
Relative mean d,eviation = 1/3(0·09 +O·ll + 0'02) X 100/49'10
= 0·15 per cent..
Observer 2. 49·40; 49·44; 49·42. Mean = 49'42 per cent.
Relative mean error = (49·42 - 49'06) /49'06 = 0·73 per cent.
_Relative ~ean deviation = 1:/3 (0·02 + 0·02 + 0'00) X 100/49·42
; 0 ·03 per cent.
The Theoretical Basis 0/ Quantitative Analysis 187
The. analyses of observer 1 were therefore accurate and precise ;.
those of observer 2 were unusually precise, but less accurate than
those of observer 1. Some small source, of constant error appears
to be present in the results of 2.
I. 73. Classification of errors.-The errors which affect an
experimental result 'may be conveniently divided into those
of the determinate and the indeterminate kind.
Determinate or constant errors. These are errors which can
be avoided, or whose magnitude can be determined. The
most important of these are:
1. Personal errors.-Thes·e are due to-some factor for which
the individual analyst is responsible; they form part of the
"personal equation" of an observer. Examples are: the
inability of some persons to judge colour changes sharply in
titrations, which may result in a slight overstepping of the
end point; the persistent use of an off-centre position in
observing the swings of a balance pointer.
2. Instrumental and reagent errors.-These arise from the
faulty construction of balances, the use of uncq,librated or
improperly calibrated weights, volumetric apparatus and'
other instruments; the attack of reagents upon glass-ware,
porcelain, etc., resulting in the introduction of foreign
materials; volatilisation of platinum at very high tempera-
tures; and the utilisation of reagents containing impurities.
3. Errors o.f method.-These originate from such factors as
incorrect sampling, incompleteness of a reaction, appreciable
solubility of a precipitate, adsorption errors, coprecipitation
and post-precipitation, and a difference between the end point
and the stoichiometric point of a reaction.
Indeterminate or accidental errors. These errors manifest
themselves by the slight variations that oecur in successive
measurements made by the same observer with the greatest care
under as nearly identical conditions as possible. They "are due
to causes over which the analyst has no control and wIiich, in
general, are so intangible that they are'incapable of analysis.
If a sufficiently large number of observations is tak~n, it can be
shown that these errors follow the law of chanc~. the equation
for which is :
. h -112)12,
y=--e
Vii
where y is the frequency of occurrence of aD error of the mag-
nitude x, and h is a constant, the value of which depends upon
the character of the measurements and which affords a
measure of their precision: 'TT and e have their usual signifi-
cance. The curve represented by this equation is called the
probability curve or the error curve; it is shown in Fig. ~1-31.
An inspection of this
curve shows: (a) small
errors occur more fre-
quently than large
ones; (b) large errors
occur relatively infre-
quently ; (c) positive
and negattve errors of
-x o the same numerical
Magnitude oj Error magnitude are equally
FIG. 1-31. likely to occur. It is
evident from (c) that the best value to select for a series of
observation.s is the arithmetical mean.
Another quantity that should be mentioned here is the
probable error (or probable deviation). By the term probable
,error is meant, not the error most likely to occur, but a quap.-
tity of such magnitude that, in a given series of observations,
it will be equally probable that there will be the same number
of errors less than it as there are greater than it, the errors
being arranged in order of their magnitude without regard to
sign. It can be shown that -the magnitude of the probable
error of a single observation is given by the formula:
P~obable error of a single observation
-- 0·6745 Jd: + d~ + .......

1 d;, h
were ddt
l' 2, e c.,
n-
refer to the deyiations from the mean, and n is the number of
determinations. The probable error of the arithmetical mean
of a whole series of determinations is given by :
Probable error of the arithmetical mean
= 0.6745 Jd~ + d~ + ..... d;
n (n - I)
= Probable error of a single determination.
y'n
In both calculations 2/3 is usually substituted for 0·6745.
The problem sometimes arises of evaluating a series of
results obtained by two or more analysts who have followed
the same method to the best-of their ability, or even different
methods', provided all obvious errors inherent in the method
have been eliminated. It is desirable for this purpose to
employ a more reasonable or conservative estimate of the
uncertainty in each analyst's result than that furnished by the
probable error or deviation. If the individual results are
available, the probable error of each analyst's mean results
should be computed and. then some reasonable: multiple of it
taken as an expression of the uncertainty in the result. No
fixed lJlultiple ot the probable error is universally employed,
but 3 and 5 are frequently used. For purpose of illustration
we'shall employ the factor 3 to denote. the reasonable uncer-
tainty in the arithmetical mean. This is then added to, or
subtracted from, the.arithmetical mean to obtain the reason-
able limits represented by the result. Sets of results so
treated are regarded as consistent if they show at least one
common value within their reasonable limits. IAn example
will make this clear. The table below contains the results
obtained by two analysts for the percentage of a constituent
A in a certain ore: the various quantities referred to in the
preceding discussion have also been calculated.
,
Probable Probable Reason.
error of error of able
Analyst Percentage of AriJb· Mean a single tbe limit of
A found metical devia- deter· aritb· Ibll
mean tion mina-
lion
metical
mean
mean
.
1 68 •
25 - - - - -
1 68·27 68·30 0'04 0·032 0·016 0·048
68'33
68·35
-
-
-
-
-
-
-
-
--
68·20 68·26 - - - - --
2
68·21
68·21
68·26
68·27
-
68·24
-
0·027
-
0·02
-
0·007 0'02
68·22 68·28 - - -- - -
68·25
.
.
-
.
- - -
The first analyst's results lead to a value of 68·30±O·048
(i.e., to limits of 68·252 and 68·348), and the second analyst's
.to 68·24 ±O·02 (i.e., to limits of 68·22 and 68·26). The two
results are consistent because they overlap within a small but
definite portion of the limit range. The results would be
reported:
1. 68·30 ± 0·05; 2. 68·24 ± 0·02,*
* For a more detailed study. see J. W. Mellor. Higber Matbematics for
Students Of Chemistry and Physics, 1929 (Longmans Green and Co.) and R . .A.
Fisher, Statjstical Methods for Research Workers, 1936 (Oliver and Boyd).
190 Quantitative Inorganic Analysts
I, 74. Minimisation of errors.-Determinate errors can
often be materially reduced by one of the following methods:
(i) Calibration of apparatus and application of correctiom.
All instruments (weights, flasks, burettes, pipettes, etc.)
should be calibrated, and the appropriate corrections applied
to the original measurements. In some cases where an
error cannot be eliminated, it is possible to apply a correction
for the effect that it produces; thus an impurity in a weighed
precipitate may be determined and its weight deducted.
(ii) Running a blank determination. This consists in
carrying out a separate determination, the sample being
omitted, under exactly the same experimental. conditions as
are employed in the actual analysis of the sample. The
object is to find out the effect of the impurities introduced
through the reagents and vessels, or to determine the exoess
of standard solution necessary to establish the end point
under the conditions· met with in the titration of the unknown
sample. A large blank correction is undesirable, because the
exact value then becomes uncertain.
(iii) Running a control determination. This consists' in
carrying'out a determination under as nearly as possible
identical experimental conditions upon a quantity of a stan-
dard substance which contains the same weight of the con-
stituent as is contained in the unknown sample. The weight
of the constituent in the unknown can then be calculated from
the relation: .
Result found for standard Weight of constituent in standard
Result found for unknown x
where x is the weight of the constituent in the unknown.
In this connexion it must be pointed out that standard
samples which have been analysed by a number of skilled
analysts are commercially availabie. These include certain
primary standards (sodium oxalate, potassium hydrogen
phthalate, arsenious oxide, and benzoic acid) and ores,
ceramic materials, irons, steels, steel-making alloys, and non-
ferrous alloys. All of these are obtainable from the U.S.'
Bureau of Standards, Department of Commerce, Washington,
D.C. Many of these are also available as the so-called
"British Chemical Standards" and are supplied by the
Bureau of Analysed Samples Ltd., 3 Wilson Street, Middles-
brough, England; these are, however, inferior to the samples
of the U.S. Buteau of Standards, but are satisfactory for most
The Thl!Oretical Basis of Quantitative Analysis 191
practical purposes. A commendable enterprise is that of the
firm of Ridsdale and Co., of 3 Wilson Street, Middlesbrough,
England, who supply what they term" Analysed Samples for
Students." These are analysed samples of irons, steels,
non-ferrous alloys, and minerals, and are comparatively
inexpensive. The results of analyses do not attain the high
accuracy of the "British Chemical Standards," but are
sufficient for use by students in the course of their training
in the methods of analysis. The.analyses do not appear on
th~ bottle, but are given on a separate sheet, which can be
retained by the teacher. They are recommended for use in
many of the analyses described in this book.
The National Physical LaboratQry, of Teddington, Middle-
sex, England, in association with the Iron and Steel Institute,
issue a limited number of "British-Standardised Steel Sam-
ples"; these comprise standard samples of steels for sulphur,
phosphorus,· carbon, and manganese respectively and a cast
iron (for Si, P, Mn and S). It is to be hoped that the range
will be considerably extended in the future.
The standard samples are useful in revealing constant
errors, for standardisation, and for comparatively rapid
testing of the accuracy of a new method of analysis. A
detailed list is given in the Appendix (Section A, 8).
(iv) Use of independent methods of analysis. In some
instances the accuracy of a result may be established by
carrying out the analysis in .an entirely different manner.
Thus iron may first be determined gravimetrically by precipi-
tation as ferric hydroxide after removing the interfering
elements, followed by ignition of the precipitate to ferric
oxide. It may then be determin_ed volumetrically by reduc-
tion to the ferrous state, and titration with a standard solu-
tion of an oxidising agent, such as potassium dichromate or
ceric sulphate. Another example that may be mentioned is
the estimation of the strength of a hydrochloric acid solution
both by titration with a standard solution of a strong base
and by precipitation and 'weighing as silver chloride. If the
results obtained by the two radically different methods are
concordant, it is highly probable that the values are correct
within small limits of error.
(v) Running of parallel determinations. These serve as a
check on the result of a single determination and indicate only
the precision of the analysis. The values obtained for con-
stituents which are present in not too sinall an amount
should not vary among themselves by more than three parts
per thousand. If larger variations are shown, the deter-
minations must be repeated until satisfactory concordance is
obtained. Duplicate, and at most. triplicate, determination
should suffice. It must be emphasised that good agreement
between duplicate and triplicate estimations does not justify
the conclusion that the result is correct; a constant error
may be present. The agreement merely snows that the
accidental errors, or variations of the determinate errors, are
the sam:, or nearly the same, in the paraliel determinations.
I. 75. Significant. figures and computations.-The term
digit denotes anyone of the ten numerals including the zero.
A significant figure is a digit which denotes the amount of the
quantity in the place in which it stands. The digit zero is a
signifioant figure except when it is the first figure in a number.
Thus in the quantities 1·2680 g. and 1·0062 g. the zero is
significant, but in the quantity 0·0025 kg. the zeros are not
significant figures; they serve only to locate the decimal point
and can be crmitted by proper choice of units, e.g., 2·5 g.
The first two numbers contain five significant figures, but
0·0025 contains only two significant figures.
Observed quantities should be noted with one uncertain
figure retained. Thus in most analyses weights are deter-
mined to the nearest tenth of a milligram, e.g., 2·1546 g.
This means that the weight is'less than 2 ·1547 g. and more than
2·1545 g. A weight of 2·150 g. would signify that it has been
determined to the nearest milligram, and that the weight is
nearer to 2·150 g. than is either 2·151 g. or 2·149 g. The
digits of a number which are needed to express the precision
of the measurement from which the numbet:. was derived are
known as significant figures.
There are a number of rules for computations with which
the student should be familiar.
1. Retain as many significant figures in a result or in any
data as will give only one uncertain figure. Thus a volume
which is known to be between 20·5 mI. and 20·7 ml. should be
'Written as 20·6 mI., but not as 20·60 ml. since the latter would
indicate that the value lies between 20·59 ml. and 20·61 ml.
Also, ii a weight, to the nearest 0·1 mg., is 5'2600 g., it should
not be written as 5·260 g. or 5·26 g., since in the latter case
an accuracy of a centigram is indicated and in the former a
milligram.
2. In rounding off quantities to the correct number of
The Theoretical Analysis 'OJ Quantitative' A naZysis 193
significant figures, add one to the last figure retained if the
following figure (which has been rejected) is 5 or over. Thus
the average of 0'26"28, 0·2623 and 0·2626 is 0·2626 (0'2625 7 ),
3. tn addition or subtraction, there should be in each
number only as many significant figure_? as there are in the
least accurately knqwn number. Thus the addition
168·11 + 7·045 + 0·6832
should be written
168'1l + 7·05 + ()'68 = 175·84.
The sum or differenc~ of two or more quantities cannot be more
precise than the quantity having the largest uncertainty.
4. In multiplic~tion or division, retain in each factor and in
the numerical result as many significant .figures as are con-
tained in the factor having the largest uncertainty. The
percentage precision of a product or quotient cannot be"
greater than the percentage precision of the least precise
factor entering into the calculation. Thus the multiplication
1·26 X 1·236 X 0·6834
should be obtained by using the values
1·26 X 1·24 X 0'68,
and the result expressed to three significant figures.
Where a large nutnber of multiplications and divisions
are to be made, the use of logarithms is recommended. Four-
figure logarithm tables (Section A, 16) are sufficiently precise
if interpolation is used; five-figure tables (Section A, 17) are
preferable as interpolation is avoided. A 10 inch slide rule
is accurate to about 0·25 per cent, ana is useful in checking
calculations. The Otis King's calculator* has an accuracy
of about.4 times that of the 10 inch slide rule, and the rotary
scales are 66 inches long; it is of convenient size for the
pocket and is very useful in the analytical laboratory .
•
• Obtainable from Carbic Limited, 51 Holborn Viaduct, London, E.C.I,
England.
H
CHAPTER II
EXPERIMENTAL TECHNIQUE OF QUANTITATIVE
AN)\LYSIS
n. 1. BALANCE, WEIGHTS, ANn. WEIGHING
The l¥1a1ytical balance, One of the most 'important tools
of. the analytical chemist .is the balance. I t is therefore
essential that fhe, student should possess a proper under-
standing of its construction, use and care. Balances differ
in details of construction and in sensitivity. For most
analytical purposes a balance with a maximum load of 100
to 200 grams and a -sensitivity of 0.1 mg. is all that is required.
The underlying principles of the theory and construction of
the analytical balance will be discussed in the following
paragraphs.
Essentially the balance may be regarded as a rigid beam or
lever Be having a central fulcrum and two arms of equ~l
-1<-" d, ... ~-+- _. :d ••• >I length; the two ends of the beam
2
carry. prism edges upon which the
8 c
balance pans ~re -supported by
means of a suitable suspension (Fig.
2-1)" Let us suppose a body having
M, M2 the mass M 1 is p1aced on the left-
p hand pan of the balance; the pointer
FIG. 2-1.
P will be deflecte<J to the right.
To restore the pointer to its
original position, bodies of known mass, termed" weights,"
are added to the right-hand pan. When equilibrium is
restored, the principle of the lever requires that the following
relation holds:
FIX d l = F? X d'l.'
where PI and F2 are the forces acting upon the left-hand aiid
right-hand prism edges respectively, and d l and d'l. are the
respective. distances of these from the central prism edge.
Since the balance has equal. arms, dl = a'l. and FI = F 2 •
Now the origin of the forces PI and P? lies in the attraction
of gravity on the bodies. in the left-hand and right-hand pans
respectively, or, otherwise expressed,
F; = MIg and F.2 = M~,
where Ml and M2 are the masses (or quantities of matter)
in the left-hand and right-hand pans respectively, and g is
194
Techniq~e of Quantitative Analysis 195
the acceleratIon due to gravity. Strictly speaking, Fl and F'I.
are the tru~ weightS in the two pans. But:
Fl M~ Ml
F2 = M 2g = M 2 '
i.e., the ratio of the forces with which the bodies in'the two
pans are attxacted by gravity is equal to the ratio of the two
masses. In quantitative analysis we are interested only in
the amount of matter in the body, i.e., in its mass: this is
independent of g. At any given place, the """eights are pro-
portional to the masses. I t has become customary to employ
the term weight" synonymously with the mass, and it is in
II
this sense that " weight" is employed in quantitative analy-

sis. The analytical balance, strictly speaking, determines
mass and not weight.
n, 2. DeScription of a typical analytical balance.-The
essential features of a 'typical analytical balance are shown
diagrammatically in Fig. 2-2. The beam, which is constructed
. Central ~nif. edge
\."i'===.f=f=====~~c::~~~tJ
·--tP<!:':±,.;r.,""'JT'"'T'"'~~"""''':f'T,,!""'TTJ-t-Adju.t,.g ,c",W
I' ,,,I' ,, I' ~ ,,I' I I

I III Enlarged
L -____o ~
5 __'10 ________
~
15 20 ~
scale
FIG. 2-2.
as rigidly and as light in weight as possible, is mounted at its
centre upon a prism-form agate knife edge, which rests upon
an agate plate (when the balance is in use) attached to the
central beam support. The two terminal agate knife edges,
fixed at equal distances from the central knife edge, each
support a suspension (stirrup) from which the pa~s are hung,
contact being made upon agate planes fixed to the stirrups.
A long pointer is attached to the centre of the beam: this
moves ov'er a scale at the foot of the pillar and serves to
indicate the deftection of the beam from the horizontal posi-
tion when the balance is in operation. The balance is levelled
with the aid of levelling screws and a plumb line suspended
trom the column; in some balances a pair of spirit levels are
provided. In the Figure the whole of the beam is divided
into 100 equal parts so that weight adjustments smaller than
10 mg. can be made readily by moving a sn:all piece of plati-
num or gold wire (the Eider) along the beam by means of tlie
rider hook and rider carriage. For the beam graduated as
shown in the Figure a 5 mg. rider is employed; in many
balances only the right half of the beam is graduated, a 10
mg. rider is then necessary. A small weight is attached to the
pointer; this serves to move the centre of gravity and thereby
alter the sensitivity of the balance. The adjusting screws at
the ends of the beam are employed for the purpose of adjusting
the equilibrium position of the pointer so that it rests at the
centre of the scale, .thereby ensuring that the beam is hori-
zontal when unloaded. In the
better type of balances, the sen-
sitivity and equilibrium position of
the poihter are adjusted by m~ans of
a screw-nut device attached to the
centre of the beam behind the rider
bar and is shown diagrammatically
FIG. 2-3.
in Fig. 2-3; the screws at the end
of the beam and on the pointer are then, of course, omit~ed.
In order to protect the knife edges fr~m undue wear and
from injury during the transfer of weights and objects, the'
balance is provided with a device whereby the moving parts
can be raised so that the knife edges and planes are very
slightly separated. This is operated by means of a large
screw attached to the base of the balance, placed either
centrally in the front of the base as shown in the Figure, or, in
some cases, in the- centre of the left-hand side of the base. In
many balances the beam arrest is combined with the pan
Technique of Quantitative Analysis 197
arrest; in others, the pan_arrest mechanism is separately
con,.trolled by a push button (not shown in the Figure).
,II, 3. The requirements of a good bblance.-These are :
(a) The balance must be accurate and give the s.ame result in
successive weighings.
This requisite _is obtained if the arms are of equal length, if
the beam is rigid and does not bend appreciably under load,
and if all the three knife edges lie in the same plane and are
parallel to one another.
(b) The balance. must be stable, that is, the beam must return
to the horizontal position after :jwinging.
This is attained by a proper adjustment of -the 'Centre of
gravity.
(c) The balance must be sensitive, that is, 0·1 mg. should be
readily detectable with average loads.
We may d-efine the sensitivity of a balance as the angular
deflection a of the peam when a known small weight is added.
I t can be shown that the angle a is determined by the excess
of weight w producing the deflect_ion· a, the length of the bal-
ance arm d, the weight of the beam W, and the distance
between the centre of gravity and the point of support of the
beam h. Expressed mathematically* :
wd
tan a = Wh'
Since a is small, we may put tan a = a. The angular deflec-

.tion of the beam is equal to the angular deflection of the
pointer, and the latter is "djrectly proportional to the number
of divisions between the two points of rest on the scale at the
foot of the beam. This leads directly to the usual definition'
of sensitivity, viz., the sensitivityf of a balance is the number
of scale divisions that the rest point (or equiHbrium point)
is displaced by an excess of weight of 1 mg.
It follows from the above expression that: (i) the longer
the balance arms, the gre~ter is the sensitivity; (ii) the lighter
the beam, the greater is the sensitivIty; and (iii) tlle smaller
the distance between the point of suspension of the beam and
the centre of gravity, the greater is the sensitivity. These
conditions for maximum sensitivity are unfortunately con-
flicting with one another. Thus long arms are incompatible
* For deduction, see. H. A. Fales, Inorganic Quantitative Analysis, 1928,
p. 81 (Century).
t Also termed sensibility and sensitiveness.
,
198 Quantitatt've Inorganic' Analysis
with minimum weight. The length of the arm is also limited
by the fact that the time of swing of the pointer, which must
not be excessive, increases with its length. H is also import-
ant that the balance beam be rigid in order that the centre of
gravity, and fhereforef' the sensitivity, should not change
app:r.eciably with the load. The centre of gravity must
always be below the central knife edge, othawise the balance
will be in neutral or unstable ~quilibriurrl. The sensitiveness
may be changed by moving the small weight attach~d to the
pointer or, in some balances, by rotation of the nut on the
s~rew attached to 'the central portion of ,the beam, which
affect the position of the centre of gravity. In general, a
balance adjusted to give a deflection of 3 to 4: scale divisions
per milligram'is·quite satisfactory for quantitative work.
(d) The period of oscillation shoulc!: be short.
This is necessary in order that weighings can be ·made
rapidly. For this reason " short~arm " balances, which have
a period of swing 9f 5 to 10 seconds, are;almost universally
employ.ed in aI).alytical practke. The so-called" long-arm"
balances, although more sensitive, have a long time of swing
{about 15 seconds) which renders weighing, a tedious process.
Sensitivity and maximum load of a balance. In an itleal
balance, free from friction and with a perfectly' rigid beam,
the sensitwity would be independent of the load. Most
balan~es, however, exhibit a decreasing sensitivity with
increasing -load, and this change of sensitivity provides a good
criterion as tQ the maximum safe load that a balance call
carry. The criterion is: no gre~ter load ·should ever be
placed upon the balance pans tfian the . load at which the
sensitivity becomes forty per cent of its maximum value (see
Section n, 6A and Fig. 2-4:). "
n, 4. Weights~ Reference mas!Jes.-The determination of
the mass of an object with an equal-arm balance necessitates
the use of a series of reference masses termed weights. For
scientific work, the international metric system ef weights
and measures is employed. The fundam.ental standard of
mass is the intern,ation81 prototype kilogram, which is a mass
of platinum-iridium alloy made in 1887 and deposited in the
International Bureau of Weights and Measures near Paris,
H represents as closely as possible the mass,of 1000 grams of
water "at the temperature of it? maximum density (4:0C) ,
Authentic copies.of the' standard are kept by the approPliate
Technique oj Quantitative Analysis 199
responsible authorities* in the various countries of the world;
these copies are employed for the comparison of secondary
standards, which are used in the calibration of weights for
scientific work. The unit of mass that is almost universally
employed in laboratory work-, however, is the gram, which
may be defined as the one thousandth part of the mass of the
international prototype kilogram.
Multiples and sub-multiples of units in the m~tric system are
indicated by t.\le following prefixes:
kilo = a thousand times; milli = a tho-q.sandth part, or
0·001 ;
hecto = a'hundred times; centi = a hundredth part, or-
0·01 ;
deca = ten times; deci = a tenth part, or 0·1.
The following weights are frequently encountered: kilo-
gram (kg.), decigram (dg.), centigram (cg.), and milligram
(mg.). . ' •
An ordinary set of analytical weights contains the follow-
ing: Gram~: lQO, 50, 20, 10, 10, 5, 2, 1, I, 1 (or 5, 2, 2, I).
Milligrams: 500, 200, 100, 100, 50, 20, 10, 10, 5, 2, 2, 1.
Weights smaller than 10 mg. are hot actually used in weigh-
ing, and are therefol'e not really epsential.. In addition to the
weights, a set includes two" riders," which are small weights
of 5 or 10 mg. denomination, l11ade of platlnum. gold, or
aluminium wire, for use on the beam of the balance. The
weights from 1 gram upwards are genera'1ly constructed of
polished brass; these are often pia ted with' gold or platinum
to avoid corrosion. Recently weights of stainless steel have
been introducec;l. The fractional weights, from 0·5 g. down-
wards, are made of platinum, gold, aluminium or tantalum;
they are usually square in shape and have one comer turned
up to facilitate han9-ling. The nominal mass is stamped on
each .weight, the numbers indicating grams, fractions of a
gram or milligrams. Duplicate and triplicate weights are
stamped with one or two asterisks. For handling the
weights, a pair of forceps, frequently tipped with ivory or
some similar material, is provided: on no account should
weights be touched with the fingers.
All weights should be calibrated (Section li, 8). The
values, may change with time, c:.nd it is advisable to recali- ,
brate the weights at least once a year.
* The National Physical Laboratory (N.P.L.) in Great Britain, the Bureau
of Standards in U.S.A., etc.
,, .
U,5 .. Care and use of the balance.-l. The balance should
be placed upon a firm foundation, which is as fr-ee from
mechanical vibration as possible. The ideal foundation is a
concrete or stone slab resting upon brick piers, which are
either sunk into the ground or, ·if this is not practicable, into
the stone or concrete flooring beneath the floor boards of the
labomtory. If this is not possible, the balance should be
placed upon a stout table or shelf and protected, when
necessary, by sbock absorbing media, such as ,cork mats or
sheet rubber under the table. It is best to keep the balance
in a room separate from the laboratory in order to protect it
from fumes. Its position should be such that the two arms
are at the same temperature. Hence the balance should not
be placed with one side near a window pr radiator, or in direct
sunlight, or in a position where it is likely to come into contact
with draughts.
The balance should be level. This adjustment may be
made with the aid of the levelling screws. •
2. When not in use, the balance beam and the pan supports
should be raised so as to protect the agate prisms and planes.
The doors of the balanc~ should be kept closed as much as
possible.
3. 'To release the balance, the beam and pahs should be·
lowered. very carefully and set swinging by fanning one pan
gently with the hand, or by carefully turning the arresting
screw. The beam should never be set swinging by touching
the pan with the' forceps or with some other object, or by
suddenly lowering.it upon its prism edges.
4. Objects to be weighed must be allowed to attain the
temperature of the balance before weighing is attempted,
otherwise the air currents produced inside the balance case
may introduce serious errors. If the objec~ has been heated,
sufficient time must be allowed for -cooTing. The tim.e re-
quired to attain the balance room temperature varies with
the size, etc., of the object, but as a rule thirty to forty minutes
is sufficient.
5. Weights and the oDject to be weighed should always be
placed in the centre of the pans.
6. During a weighing or upon any occasion when objects
are being added to or removed from the pans, both the. beam
arrests and the r:'.n supports must be raised so as to protect
the prism-edges from injury. Always close the balance case
when starting to use the rider.
.. 7. No chemicals or objects which might injure the balance
pans should ever be placed directly upon them. Subst'ances
must be weighed upon suitable containers, such as watch
glasses, small beakers, weighing bottles gr crucibles. Liquids
and volatile or hygroscopic solids must be weighed in a
- tightly closed vessel, such as a stoppered weighing bottle.
S. The balance must not be overloaded (see Section n,
3, d).
9. Weights should be handled only with the forceps 'pro-
vided with the box. .
10. Nothing must be left on the pans when the weighing
has been completed. If any substance is spilled accidentally
upon the pans or upon the floor of the balance case, it must be
removed at once. The pans should be lightly brushed perio-
dically with a camel-hair brush to remove dust which may
have collected.
11 ... If the balance requires adjustment, the teacher should
be notified immediately. J
II, 6. METHODS OF WEIGHING.

There are three methods of weighing, viz.,
(A) Direct weighing.
(B) 'Gauss's methods of double weighing.
(C) Borda's method of substitution.
Of these, method A, in one of its various modifications, is
most widely employed in quantitative analysis. It will
therefore be discussed in some'detail.
(A) Direct weighing. The object is placed on the left.
hand balance pan and weights are add.ed to the other p'an until
the beam returns to its original equilibrium position. The
modifications of th,is method differ only in the procedu:r;e for
determining the equilibrium position of the beam.
(i) Weighing. by point of rest and sensitivity. 1he first step
in any method of weighing is the determination of the equili-
brium position of the beam when the pans are unloaded. The
position of the beam is of course indicated by the po.sition of
the pointer .on the scale. We shall call the equilibrium posi-
tion the zero point when the pans are empty, and the point of
rest when the pans are' loaded.
Determination of the zero point. See that the balance is'
level; if necessary, turn the levelling screws at the base until
the balance is suitably adjusted (spirit level or plumb-line).
Note also whether the prism edges are in proper position with
H*
202 QUa'ntitative Inorganic A nalysz's
resp~ct to their bearings, and also wheth~r the pans a;e free
from dust (the latter can be removed, if necessary, by means
of a camel-hair brush). The rider should be placed on the
zero graduation if the whoJe ,of the beam is graduated; if
only the right half of the beam is grttduated, the rider should
b~ removed from the. beam. The beam and pan arrests are'
lowered very carefully, and the oeam set swinging either by
fanning one· pan gently with the hand or, more usually, by
careful manipulation of the arresting control. The doors of
the ,balance are then closed.
The pointer scale may be regarded as being graduated as
shown in Fig. 2-2, i.e., numbered from zero to twenty. It is
convenient to have the pointer swinging about 5 to 6.divisions
on either'side of the centre mark (10) of the scale. The beam
is allowed to make two or three swing~ and then readings are
taken to the nearest tenth. of a division of the extreme posi-
tions of the pointer. An odd number of swings, usually three,
is taKen on the left-hand side and even number, usuall:? two,
on the right-hand side. The eye of" the observer should
remain in the same position to avoid errors due to parallax.
The two sets of observations are averaged separately, and the
mean of their~average taken. This mean value is the zero
point.* 1he procedure is best made clear with the aid of an
example. Five ~uccessive readings of the poin\er might be :
L~ R~M
(1) 5·3 (i) 15·2
(ii) 5·7 (ii) 14·9
(iii) 6·0 Average '15'1
Average 5·7
. Mean = (5·7 + 15·1) ( 2 = 10'4 = zero point.
After s..ome practice, the·zero point may be calculated from
three successive readings of the extreme positions instead of
five. If -the. zero point lies outside the positions 9 and 11,
i.e., more than one division from the centre line of the scale,
the student should ask the teacher to make ali adjustment on
thel nuts attached to the ends • of tnt beam or, for some bal-
ances, on the nuts upon a small lever attached to the centre of
the beam.
* For a mathemati~al justification of this procedure, see H. A. Fales,
- Inorganic Quantitative .A nalysls, 1928, p. 85. It must be realised that,. the
beam swings about the prism edge like a compound pendulUII\ and that
the oscillatlOtl~ are damped, or shortened, by the friction in th~ bearings and
by air resistance; consequently each swing will be slightly less in amplitude
than the preceding one.
Technique
\
of Quantitative Analysis 203
The next step is the determination of the sensitivity of the
balance.
Dettrmination of the sensitivity of the balance. The' sensi-
tivity of a balance depends inter alia upon the distance be-
tween the centre of gravity and. the point of suspension of the
beam (Sectioh U, 3). Since most beams are not perfectly
rigid, the distance, and consequently the sensitivity, will
change with the load. Whether the sensitivity increases or
decreases with increase of load depends upon the relative
positions of the beam prism-edge and the pan prism-edges.
The sensitivity has already been arbitrarily defined as the
nqmber of scale divisions that the point of m~t is displaced by
an excess of weight of one milligram.*
The pans and beam are,arrested and the rider is placed
exactly on the 1 mg. gra~uation of the beam. The beam is
then- set swinging and the rest point determined exactly as
described above. ;rhus, if the zero.poiI,lt is 10·4 and the rest
point with a load of 1 mg. is 6'3, tht:n the sensitivity at this
load is 1 mg. = 4'1 s<;ale divisions, and the displacement of
one division corresponds to 0'0010/4'1 = 0·0002 g. The
-sensitivity for a load of 10...grams is obtained by.. placing 10 g.
in each pan and
determining the rest
.
4
point; the rider is
displaced 1 mg. and ~
the rest point again ~ 3
determined. The ~
~
......... ,..........
r--...... ....... ..._
difference gives the :! .__
h--
sensitivity. Similar c:. 2
estimations are made ~ •
for loads on each pan ~
of '30, 40, 50, and ~ 1
100 g. The results
may then be plotted
as shown in Fig: 2-4. 0 10 20 30 40 50 60 70 80 90 100
The sensitivity at Load (grams) in each pan
any load can then FIG. 2-4.
befound by reference Sensitivity Curve of a balance.
to the curve. Alternatively, a table may be constructed
showing the sensitivity at various loads, and also the number
of mg. corresponding to one scale division. Either the
* An alternative! definition which is someti!Ues employed is the following ~
the sensitivity is the weight necessary to shift the equilibrium position of the
pointer by one scale division.
204 Quantitative Inorganic A naz'ysis
graph or the table should be placed in a conspicuous position
in the balance case.
Weighing by the method of swings. Having determined the
zero point and the sensitivity at various loads, the actual
weighing operation may be performed. The object is placed
on the left-hand pan and weights are added to the right-hand
pan in systematic order. The beam and pans should be raised
before an object or a weight is placed on the pans. A weight
estimated to be slightly heavier than the object is removed
from the box with the forceps and placed in the centre of the
right-hand pan. The beam is lowered just enough to see
which way the pointer moves; the beam is theIf raised. If
the weight is too heavy, it is removed and the .next smaller
weight substituted for it. If the-latter is too light, it is
allowed to remain on the pan and the next lower weight added.
This process is repeated in a systematic order until the gram
weight is reached .• The. fractional ~eights are similarly
added in order of "decreasing denomination down to and
including the 10 mg., weight. Here, particularly in the con-
cluding stages, it will be necessary to lower the beam and pans
completely in-order to detect the inequality.
The final adjustment is made with the rider with the balance
case closed. For very accurate weighing, the rider is adjusted
to the nearest mg. and the tenths (or hundrMths) of a milli-
gram are then calculated from the difference between the zero
point and the rest point, the sensitivity being known 'for the
load on the pans. An example will make the method of
ca!culation clear. A certain object is found to weigh between
12·673 g. and 12.·674 g. The rider is then placed exactly on
division 3 (mg.) and t~ rest point determineq. by the method
of swings. Suppose this is 12·4. The zero point is, say,
10·4. The sensitivity of the balance, estimated from 'the
sensitivity-load graph, is 3·8 divisions per mg. The weight
is therefore:
" .
12·673 + (12·4 - 10·4) X 0'001/3'8 = 12·673 + 0·00053
= 12·67353 g. or 12·6735 g. (to the nearest 0·1 mg.).
(ii) Weighing by the method of short swings. For ordinary'
analytic~l work, entirely satisfactory weighings to an accuracy
of 0·1 mg., can be made by the method of short swings. The
~wings observed are so short that the damping effect produced
in two successive swings is inappreciable. The zero or rest
point is determined by adding together the observed values of
two successive swings on either side of the centre of the
pointer scale, and taking the mean. The swings may vary
from 4 divisions on the pointer scale to tile shortest one that
is distinctly visible; those from 1 to 2 divisions on' either
side of the centre of the scale are usually satisfactory. As
an example, we -may consider two successive swings + 2·2
and - 2·0; the centre of the scale is here taken as the
reference point, values to, the right being positive and those
to the left negative. The zero point is (+ 2·2 + - 2·0)/2 =
+ 0·1. This method of weighing is recommended for flU
routine work, but is not suitable for the calibration of weights
(see Section n, 8).
(iii) Weighing of chemical samples. It will be convenient
at this stage to discuss the application of the two methods of
direct weighing just described to the weighing of chemical
samples. Most chemicals 'are weighed by differenc~ in order
to avoid exposing the sample to the atmosphere, from which
dry substances may absorb moisture in varying amounts.
If the sample has been dried, it should be pla'Ced in a weighing
bottle and kept in a desiccator, containing a suitable drying
agent, until required for weighing. The most convenient
form of weighing bottle is that provJded with a ground-on
cap (Fig. 2-5a). That with a ground-in stopper is not recom-
mended parti<;ularly for weighing powpers; there is al~ays
I o(---3·25x---'" I
.. '" - - - - - - - - - - 1'X - - .)oJ
(a) (b)
•
FIG. 2-5.
a danger that the small particles may adhere to the ground

surface .and that when the stopper is pressed into place, they
may either tend "to cause the stopper to stick or else be forced
out and lost. A special type of "duplex" weighing bottle,
together with the relative dimensions, is shown in Fig. 2-5b.
This is fitted with externally ground caps, and possesses
many obvious aq,vantages. Among these may be mentioned
the flattened side which gives great stability, the wide mouth
which permits facile addition and removal of solids, the nar-
row mouth which allo\ys of easy transfer of solids or liquids
into narrow-neck receptacles, and finally the designed slopes
which permit easy access of brush or stream of liquid into
and against all parts of the interior.
To weigh a substance by difference, weigh a glass-stoppered
weighing bottle (Fig. 2-5a) containing the sample. Approxi-
m~tely the desired amount is then transf~rred into a suitable
container (beaker, flask; etc.) by appropriate shaking, and the
bottle reweighed. If desired, the sam pIe may be removed
with the aid of a small clean spatula (which should preferably
be of platinum) which is kept in the bottle. Sometimes the
sample is removed with the aid of a clean spatula, spoon or
scoop, which is not weighed; it is then of the greatest im-
portance that. every, particle removed from the bottle be
transferred to the container. The spatula or scoop should be
carefully brushed with a small camel-hair brush.
In the actual transfer' of the weights from the box to the
balance pan, it is desirable to place the gram weights in the
centre of the pan, the .decigram weights to one side 'and the
centigram pieces to the other side. This will assist in reduc-
ing the likelihood of error in recording the weighing. The
weig1J.ts on the pan afe noted and their values' oonfirmed by
observing the vacant p'laces in the weight box. As a final
check, the value of each weight as it is returned to the box is
noted. All weight~ should be recorded in ink in a note book
at the time that they are made: loose sheets of paper should
never be used as they are. readily lost.
If the sample is unaffected by exposure to the air, it may be-
weighed on a watch glass: A clean, dry wat,ch glass is
weighed; portions of the sample, which' should preferably be
in the form of a powder, are added by means of a spatula or
scoop until the weight is approximately correc~, and the
combined weight is then determined. It ·is often convenient
to have a pair of counterpqised watch glasses for this purpose,
i.e., a pair of watch glasses which weigh so nearly the same
that they can be balanced with the aid of the rider alone.
One watch..glass is placed on each pan, and weights art placeq
on the right-hand pan amounting to the desired quantity of
the sample; the latter is th~ri added with a spatula or similar
apparatus.
Liquids must be weighed in closed weighing bottles. The
.' most convenient method is to weigh the empty bottle and
J
Technique of Quantitative Analysis - 207
stopper: the approximate amount of liquid is introduced by a
graduated pipette or glass tube, and the bottle is reweighed. The
entire contents of the bottle is then washed into the second
container.
Sometimes ignited precipitates are hygroscopic or tend to •
absorb carbon dioxide from the air. Weighing must then be
made as rapidly as possible. These precipitates are ignited'
to constant weight; the approximate weight is therefore
known from a previous ignition. In the final weighing it is
perhaps best to adjust the rider to the nearest milligram and
then to determine tenths ,of a milligram by the method of
swings. . •
An experimental point which merits attention is that a
vessel should always be weighed filled with the same gas.
Thus in the determipation of carbon by combustion in
oxygen and the absorption of the resultant carbon dioxide in
a U-tube containing soda lime or some other substance, it is
important to weigh the absorptiop tube filled with air both
before and after the determination. To get some idea of the
error resulting from the neglect of this precautioR, let us take
the case of a U-tube of, say, 50 ml. capacity. One litre of air
at 20°C and 760 mm. pressure weighs approximately 1·20 g.,
and one litre of oxygen under similar conditions weighs 1·33 g.
, The difference in weight produced by sUbstituting oxygen for
air will be (1·33 - 1'20) X 50/1000, or 0·0065 g. Since
the weight of carbon dioxide absorbed, is usually of the order
of 0·1 g., the neglect of this precaution will introduce an error
of about 6·5 per cent. '
A common procedure in quantitative analysis is to dissolve
a known weight of a substance in a definite volume of liquid,
which is usually distilled water or distilled ~ater to which -a
reagent has been added. A graduated flask is generally
employed for this purpose. The operation is carried out as
follows. Insert a short-stemmed funner into the neck of the
flask. Transfer the requIred amount of substance from the
weighing bottle to the funnel, taking care that no p'articles are
lost. After the weighing bottle has been reweighed, 'the
substance in the funnel is washed down with a stream of the
liquid. The funnel i~thoroughly washed, inside and out, and
then removed from fhe flask; the contents of the flask ~.re
dissolved, if necessary, by shaking or swirling the liqu'id, and
then made up to the mark, if an exac;t volume is required ...
If a w~tch glass is employed for weighing out the sample, the
contents are transferred as completely as possible .to the.
208 Q1lantit~tive Inorganic ,A nalysis
funnel; the last traces remaining on the watch glass are
transferred by washing. The procedure is then as above.
B. Gauss's method of double weighing. This method will
eliminate any errors due to the ineR_uality of the arms. Let
us suppose that the lengths of the arms are d 1 and d'}, respec-
tively (see Fig. 2-1),and the true mass of the object is M t .
If the object is placed in the left-hand pan and is counter-
balanced by weights Wr on the tight-h_and paQ., we have by
the principle of moments: .
M t X d 1 = Wr X d2 (i).
If now the object is pla<1ed in the right-haIyj pan and is
counterbalanced by weights WI on the left-hand pan, we have:
M t X d2 = W, X rJ,1 (ii).
Upon multiplying these two expressions together and can-
celling out the common factor dId z, we obtain:
M~ = WrW". or M t = v'WrW, (iii),
i.e., the mass is equal to the square root of the product of the
two apparent weights. Since d 1 and d 2 are yery'nearly equal;
it is sufficiently accurate to write:
M t = (Wr + W,) /2 (iii'),
i.e., the true mass i~ equal to the arithmetic ~ean of the
apparent weights. _
C. Borda's method of weighing. This is another method
for determining the true mass of an object; it is independent
of the inequality of the balance arms. The object is placed
on the right-hand pan, and counterbalanced or tarOO by any
Gonvenient counterpoise, which may be the weights of a
duplicate set, lead shot or sand. The rest point of the balance
is determined. The object is then removed and replaced by
weights which give "the same rest point. The mass of the
object is thus equal to the mass of the weights on the left-
hand pan. This method is useful for weighing large objeCts.
It "is also employed in the calibration of weights, but the
precision is not so great as in Gauss's method.
II. 7. Errors in weighing.-The chief sources of error are
the foHowii)g :
1. Change of the condition of the containing vessel or of the
substance between successive weighings.
2. Inequality of the lengths of the 15alance arms.
Technique oj Quantitative Analysis ro9
. 3. Effect of the buoyancy of the air upon the object and
the weights. .
4. Inaccuracy of the weights.
- The first source of error is occasioneq by the change in
weight of the containing vessel (a) by absorption or loss of
moisture, (b) by electrification of the surface caused by rub-
bing, and (c) by its temperature beingdifferentfrom that o£the
balance case. These errors may be largely eliminated by
wiping the vessel gently with a linen cloth) and allowing it to
stand at least 30 minutes in the balance room before weighing.
The electrification, which may cause a comparatively large error
particularly ij both the atmosphere and the cloth are dry, is
slowly dissipateq. on standing, but is best remove.d by placing
a piece of pitchblende or other radioactive material in the
balance case to ionise the air: The best method, if practic-'
able, is to use a vesse1 of similar dimensions as a tare. Hygro-
scopic, efflorescent, and volatile substances must be weighed
in completely closed vessels. Substances which have been
heated in an air oven or ignited in a crucible must be allowed
to cool in a desiccator containing a suitable drying agent.
The time of cooling in a desiccato! cannot be exactly speci-
fied, since it will. depend upon the temperat~re and the size
of the crucible as well as upon the material of.which it is com-
posed. Platinum vessels require a shorter time than those
of porcelain, glass or silica. As a general rule, platinum
crucibles should be left in the desiccator for 25 minutes, and
cruciQles of other 1l).a..terials for 35 to 40 minutes before being
weighed.
The use of drying agents or absqrbents for carbon dioxide
in the balance case is generally of little value except where
atmospheric conditions are such as to otherwise cause serious
corrosion' of the balance. These reagents give comparatively
little protection during weighing, ftr the small surface of
absorbent usually exposed caJ.1.not quickly purify the air
introduc~d when the case is opened. However, if reagents
are to be used, soda-lime and fused calcium chloride ar~
satisfactory. .
The error due to the inequality of the lengths of the balance
arms may be. completely eliminated by Gauss's method of
double weighing (Section II, 6B) or by Borda's method of
substitution (Section II, .aC). In a good analytic'al balance
the arms will seldom differ in length by more than 1 part in
50,000 so that the error introduced by the direct method of
weighing is not" large. The error is still further reduced in
2""10 Quantitative Inorganic Analysis
routine work in quantitative analysis where the sample and
the ignited or dried precipitate is weighed on the same
balance and on the same pan. Here absolute weigp,ts are not
essential, since it i~ only the ratio of the weights with which
we are concerned. It is cle~r that if both the sample and the
precipitate are both affected by the same small error, the
ratio 9f the weights will not be, appreciably affected. Where
absolute weights are required, as in the calibration of weights
or of volumetric apparatus, one of the two methods mentioned
at the beginning of this paragraph must be used.
Effect of the buoyancy of the air. Reduction of weighings to..
vacuo. When a substance is immersed in a fluid,. its true
weight is diminished by' the weight of the fluid which it
. displaces. If the object and the weights bp.ve the same
density, and consequently the same volume, no error will be
introduced on this account. If, however,' as is usually the
case, the density of the object is different from .that of the
weights, the volumes of air displaced by each will be different.
If the substance has a lower density than the brass weight~,
as is usual in analysis, the former will displace a greater
volume of air than the latter and it will therefore weigh less
in air than in a vacuum. Conversely, if a denser material
(e.g., one of the precious metals) is weighed, th~ weight in a
vacuum will be less than in air. r.
Let us enquire into the magnitude of the error. First, let
'l1S consider the weighin_g of a litre of water, first" in vacuo,"
and then in air. It is assumed that the flask containing the
water is tared by an exactly similar flask, that the temperature
of the air is 20°C and th~ barometric pressure is 760 mm. of
mercury. The weight of 1 litre of water" in vacuo" at
20°C and 760 mm. is 998·23 grams. If the water is weighed
in air, it will be found tbat 998·23 grams are too heavy. We
can readily calculate the difference. The weight of 1 litre
of air displaced by the water is 1·20 graxps. The brass
weights have a density of 8'4, hence they will displace
1)98'23/8'4 = 116·9 mI., or 116·9 X.l·20/1000 = 0·14 grams
01 air. The net difference in weight will therefore be 1·20 -
0·14 = .1'06 grams. Hence the weight in air of 1 litre of
water under the experimental conditions named is 998·23 -
1·06 = '007 ·17 grams" a difference of 0·1 per cent from the
weight" in vacuo."
Let us now extend our enquiry to the case of a solid, such
as potassium chloride, under the above conditions. The
Technique of Quantitative Analysis ·211
density of potassium chloride is 1·99. If 2 grams of the salt
are weighed, the apparent loss in weight (= weight of air
displaced) is 2 X 0'0012/1'99 = 0·0012 grams. The apparent
loss, in weight for the brass weights is 2 X 0'0012/8·4 =
0·00029 grams. Hence 2 grams of potas'sium chloride' will
weigh 0·0012 - 0·00029 = 0·00091 grams less in air than
" in vacuo," a'difference of 0-05 per cent. '. •
It must be pointed out that for most analytical pu{"pose~
where it is desired to express the results' in the form of a
percentage, the ratio of the weights in a~r, so far as solids are
~concerned, will give a result which is practically the same as
that which would be given by the weights" in vacuo." Hence
,.no buoyancx correction i$ necessary in these' cases. How-
ever, where absolute weights are required, as iIf the caliocation
of volumetric g1assware, corrections for the buoyancy of the.
air must be made (compare Sectiori n. 20).
Let us now consider the general case. I t can be shown that
if Wv = weight "in vacuo," Wa = apparent weight in air,
da = density of air, d w = density of the weights, and db 1=
.density of the body, then: I
Wv Wa)
Wv = W a + da ( db - d w (i) .
The density of the air will. depend upon the humidity, the
'temperature and the pressure. For an aVQrage humidity
(50 per cent) and average conditions of temperature and
pressure in a laboratory, the density of the air will rarely fall
outside the limits o·oon and 0-0013 g./ml. ~t is therefore
permissible for analytical purposes to take·the weight of 1 ml.
of air as Q'0012 g.
Since the difference betwee~ W v and Wa does not usuap.y
exceed 1 to 2 parts per thousand, we may write:
Wa Wa)
Wv=Wa+da ( d ;:-"- d ' •
b w
= + Wa { 0·0012 (~b -
Wa :'4)} [for ~:ass ;.:~ghts,
= Wa + kWa/l000,
, where k = 1·20 (_!_ --.! ).

db S·4
The values of k for Cia = 0·0012 and d w = 8·4 have been
calculated and are collected in the following Table. If a
21'2 ~ Quantitative Inorganic Analysis
substance of density db weights Wa grams in air, th~n
Wa.k milligrams are to be added to the weight in air in order
to obtain the weight" in vacuo." The correction is positive
if the substance has a density lower than that of brass and
negative if the density of the substance is greater than that of
brass.
.
TABLE XXIV. REDUCTIONS OF WEIGHINGS MADE IN AIR WITH
BRASS WEIGHTS TO "VACUO."
db k db k db k
0·5 + 2·26 1·9 + {)·49 n·o .. - 0·03

0·6 + 1·86" 2·0 + 0·46 12·0 - 0·04
0·7 + 1·57 2·5 + 0·34 13·0 - 0·05
0·8 + 1·36 3·0 + 0·26 14·0 - 0·06
0·9 + 1·19 3·5 + 0·20 15·0 - 0·06
1·0 +- 1·06 4·0 + 0·16 16·0 - 0·07
~ 1·1 + 0·95 4·5 + 0·13 17·0 - 0·07
1·2 + 0'86" 5·0 + 0·10 18·0 - 0,08
1·3 + 0·78 5·5 + 0·08 19·0 - 0·08
1·4 + 0·72 6·0 + 0·06 20·0 - 0·08
1·5 + 0·66 7·0 + 0·03 21·0 - 0·09
1·6 1 + 0·61 8·0 + 0·01 22·0 I - 0·09
1·7 + 0·56 9·0 - 0·01 23·0 - 0·09
1·8 + b'52 10·0 - 0"02 24·0 - 0'09
I ~
IT, 8. Calibration of weights.-All weights should be cali-

brated in order to detect any inaccuracies amongst the
weights th'emselves~ This calibrati9n should be repeated at
least once a year, because changes may occur owing to wear,
corrosion or other causes. In all good sets of weights the
duplicate and triplicate weights are distinguished by the
marking of one * or two ** asterisks on the surface. If this
is not the case, the set should either be returned to the makers
or the markings made with a suitable punch by a qualifte9
mechanic in the instrument workshop of the Institution.
The simplest .and best method of calibration is to check
each piece of a set against a corresponding piece of a set
calibrated by a nationally recognised laboratory. * 'the
method of substitution is employed. A third or subsidiary
... For the British Empire, this is the National Physical Laboratory; for
U.S.A., the Bureau of Standards, etc.
set of weights is also requireq. Fi~st. place the smallest
weight of the set to be calibrated on the right-hand pan and
balance it by the con:esponding piece of the subsidiary box 6f
weights placed on the left-hand pan. Ascertain the rest
point by the method of swings. Remove the w!!ight from the
right-hand pan and replace it by a piece of the same denomina-
tion from the standard set of weights, and determine the rest
point again. The error in the weight can :then be computed
from the difference in the rest point in the two weighings and
the known sensitivify of the balance for the load in question.
This process is repeated for every piece in the box of weights.
Finally a table is constructed showing the absolute error for
each weight. t
Complete sets 'Of standard weights are not always available.
A modification of the method due t6 T. W. Richards (1900),
which is essentially a calibration of the weights relative to one
another, is then employed. Here also the substitution method
of weighing is utilised ;..an extra set of weights, to act as ~
tare, is also required. The principle of the method is to
assume that the smallest weight in the·set, say one of the 10
mg. pieces, is correct, and the value of every other weight in
the set is obtained in terms of it. The weighings must be
made by the method of swings. The values for the larger
weights are found to differ considerably from the nominal
values because such a small standard was used. '1 t ·is there-
fore usual tQ change the relative standard to one of the larger
weights, say, to a nominal 10 g. piece, and to express all the
other pieces in terms of the larger standard. The differences
will then"be much-smaller. If the absolute value of the new
standard can be determined by comparison with a similar
standard weight (calibra'ted by the N.P.L., Bureau of Stan-
dirds, etc.), we can then readily calculate the absolute values
of all the other weights, since the ratio of the 10 g. weight .to
all the others ill the set has been determined.
The detailed procedure is as follows. The sensitivity.of the
balance at different loads must be known. If this is not the
case, it must be determined as described in Section 11, 6A.
Leave the weights for at least 30 minutes before commencing
the calibration. Place one of the 10 mg. weights, temporarily
assumed as exactly 10 mg., on the right-hand scale pan, and
counterpoise it with the corresponding weight from the sub-
sidiary...see·on the left-hand pan. Determine~h~ rest point
by the method of swings; this should be near the centre of
the scale (division HJ in Fig. 2-2). Repeat the determination
214 QUa"ntitative Inorganic Analysis
of the rest point for the. other two 10 mg. weights (0'01* and
0'01 **), retaining the original tare on the left-hand pan.
A'hy difference in their weights can be calculated from the
known sensitivity of, the balance. Next place the two 0·01
g. pieces (say, 0'01 and 0'01*) on the right-hand pan, and
counterbalance them with 'a 0·02 g. weight from the extra set.
Determine the rest poip.t. Repeat with the 0'02 g. weight
from the set, using the same tare. The calibration correction
for 'the 0·02 g. weight can then be determined, knowing that
for the 0·01 g. and 0'01* g. Continue the' comparison until
the complete set is calibrated. The following comparisons
will be made: 0'01* g. against 0·01 g., 0'01** g. against
0·01 *.g. ;" 0·02 g, against (0·01 + O·O~ *) g.; 0·05 g. against
(0'02, 0'01, 0'01*, 0'01**) g. = E 0·05; 0·1 g. against
(0'05 + E 0·05) = E 0'1 g.; 0'1* g. against ,0·1 g.; 0·2 g.
against (0'1 + 0'1*) g.; 0'5 g. against (0·2 + O·l + 0'1* +
E 0'1) g. = .E 0'5; 1 g. against (0·5 + E 0'5) g.; 1* g.
against 1 g.; 1** g. against 1* g.; 2 g. against (1 + 1*)
g.; 5 g. against (2 + 1 + 1* + 1**) g.; 10 g. 'against
(5 + 2 + 1 + 1* +. 1**) g.; 10* g, against 10 g.; 20 g.
against (10 + 10*) g.; 50 g. against (20 + 10 + 10* +.5
+ 2 + 1 + 1 * + 1**) g.
It is generally advisable to calibrate the rider, also. Let us.
suppose this is of 10 mg. deI,lomination. The Jjder is placed
on position;10 of the beam and comparison is made 'directly
with the 0·01 g., the assumed temporary standard placed on
the left-hand pan. In SOD1e sets of weights there are-only
two 10 mg. pieces; in such cases the rider is kept at position
10 during the. calibration.
The results should be tabulated as shOwn in Table XXV.
This also iIJustrates the exact method of computation. The
method of calculation of the relative values taking the 0-01 g.
as correct is obvious. Column 7, the ideal calibration value,
is' most simply obtained ,by assuming !hat tRe 10 g. weight
actua}ly possesses the weight of 10·03'85.g. (se~ Table) and
taking this as the permanent standard; there is therefore no
correction for this weight. The 5 g. pieces in this basis
shouid"weigh 5 X 10'0385/10 = 5·0193 g.; the other. ideal
calibration values are calculated in the same manner. The
differences betw.een columns 6 and '7 give the corrections to be
added or subtracted to the nominal values of the weights. A
table should ~e constructed showing the correctioIl: to be
applied to each weight: the table may also contain the
corrections for various combinations of' weights.
.Technique of Quantitative Analysis 215
TABLE XXV. CALIBRATION OF A SET OF ANALYTICAL WEIGHTS BY THE
METHOD OF SUBSTITUTION.
Tare Sensi- Differ- Relatzve Ideal Corree-

on left Nominal wi. Rest tivity etlce values iJ calibration in
pan on right pan point div. in g. 0·01 g_ lion _mg.
mg. , correct value
--
o.m\ i
0-01
0-01*
0'01**
10-4
10·4
10·4
5·0 0-0100
0·0100
0-0100
0·0100
0·0100
0·0100
0·0
0-0
0-0
-0'02- 0-01 + 0'01* 10·9 (0·"0200)
.0·02 9·9 + 0'0002 0·0202 0·0201 + 0·1
0-05 0·02, 0 '01,
0'01*,0'01** 10·7 (0'0502)
0·05 I
--
10·3 + O-OOOL 0-5003 0·0502 + 0·1
0-1 o·on + EO·05 10·6 (0'1003)
0-1 10-6 + 0·0000 0·1003 0·1004 - 0·1
0-1* 9-6 + 0·0002 0·1005 0·1004 + 0·1
--
0·2 0·1 + 0·1* 9-2 (0·2008)
0·2 10·7 - 0·0003 0·2005 0'!008 - 0·3
0·5 EO'S 11·2 (0'5016)
0-5 10·3 + 0·0002 0·5018 0'5019 - 0·1
--
I 0-5 + EO'5 11-3 (1'0034) .
1 8-3 + 0-0006 1-0040 1·0039 + 0·1
1* 9-3 + 0·0004 1·0038 1·0039 - 0·1
1*· 10·3 + 0·0002 1·0036 1·0039 - 0·3 •
-2 - I, 1*
• 10· ..
.,.
(2'0078)
2 10·9 - 0·0001 2·0077 2·0077 0·0
5 2, l,oJ.., 1*· 8·0 (5'0191)
5 9·2 - 0·0002 5-0189 5·0193 - 0·4
--.
1.0 5,2,1,1*,1*· 11·6 4·6 (lO·0380)
10 9·3 + 0·0005 10·0385 10·0385 ± 0-0
10· 11·6 0·0000 10'0380 10-0385 - 0·5 •
.--- --
20 lO, 10* 11·4 4·2 (20-0765)
20 11·4 0·0000 20-076'5 20·0770 - 0·5
I
--
50 20, }O, 10*, ri, 2
1,1*,1** 12'0 3·6 (50'1910)
50 8·5 . + 0'0010 5O'I~20 50-1925 - 0·5
Attention is directed to the fact that fQr quantitative work

• we rarely require to know the absolute weight, .since we are
nearly always concerned with weight ratios. It is 'of the
utmost i'mportance that th~ weights should agre~ among
themselves, e.g., the 2 g. piece should be exactly twice the I g.
216 Quantitative Inorganic Analysis •
weight, etc. Except for the most expensive sets, we rarely
find that these adjustments have been carried out with
sufficient accuracy. Hence it is imperative that all sets of
weights should be calibra.ted or standardised by one of the
methods just outlined.
II, 9. Other types of balances.-The instrument described
in Section II, ~ is spoken of as an ordinary balance, and is the
type that is most widely used for analytical work. In the
last few decades- <J. number of changes have been made in the
design of balances, chiefly with the objects of rendering the
operation of weighing more convenient and more rapid.
Some of these will be mentioned below.
A.' The chainomatic balahce. This balance is similar to an
ordinary chemical balance in general construction, but differs
from it in that the use
of fractional weights
below 100 mg. and the
rider are y n t ire 1y
eliminated. A rolled-
gold or gold chain is
attached to one end of
the beam and at the
other end to a small
vernier movable along
a vertical col u m n ,
/ graduated in mms. and
",.,_".,' ..... -1
8,./f.',,,,,,_ situated to the right
'fO .... .,...,
of the balance beam.
This sliding vernier or
block is controlled by
turning a handle or
s c r e w 0 u t sid e the
balance case. I t will
be evident from Fig.
2-6* that the weight of
FIG. 2-6. the chain is carried
both by the ..beam and
the vernier block; the higher the latter, the smaller is
the proportion of the weight carried by the beam. The
column is graduated to read directly in mg. up to 100 mg.,
and with the aid of th~ vernier block readings can. easily be
* This Figure is reprodl:!ced by courtesy of L. Oertling, Ltd., 65 Holborn
Viliduct, Londo~, E.C.l, England.
made to 0·1 mg. (see Figure 2-6). In some types of chaino-
matic balances, the beam is serrated into 10 equal parts from
" 0 " on the extreme left to" 10" on the extreme right, and is
aajusted with the aid of a special 5 decigram rider at zero,
each serration or notch being equivalent to one decigram.
In this type of. balance the use of fractional weights is com-
pletely eliminated. • .
The chief adv'antages of chainomatic balances are: (i)
Weights less than one decigram (in some balances one gram)
are unnecessary. (ii) .The frequent opening and closing of the
balance case towards the end of a weighing are avoided; the
chainomatic mechanism can be operated whilst the balance
is swinging. (iii) The weighings can be carri~d out rapidly.
B. The air damped balance. An oscillating system, such
as a balance, is said to be damped if it is subject to some force
which restrains its motion and decreas.es the amplitude of its
swings. All balances are damped to a slight extent since
friction cannot be entirely eliminated. In the air damped
balance, large light pistbns attached to the pans or to the
beam move loosely in large cylinders. The resultant damping
is so great that the pointer swings slowly from the centre to
the rest point within 10 to. 15 seconds. The sensitivity of the
balance is not decreased by damping. In weighing it is
therefore only necessary to note the poin't of rest. Some
balances (e.g., those of Bunge) have a scale which is graduated
to correspond to milligrams and the rest point is observed
through a microscope. In other balances (e.g., those of
Oertling) a finely-divided graticule is attached to the end of
the pointer and an enlarged refiectipl1 of the scale isprojected
on to an illuminated ground-glass screen situated 'in a con-
venient position at the top of the balance case. The graticule is
divided into 500 equat'parts, and balances are available in which
the'500 divi~ions correspond to either 0'05gram orO'5gram; in
t\1e former case weights below 0'Q5 gram are unnecessary, and
in the latter case all fractional weights are eliminated:
C. The magnetically damped balance. Damping can also
be effected by suitably suspending a small aluminium plate
from the right-hand stirrup of the balance; a counterpoise,
usually of brass is suspended from the 'left-hand stirrup;
The aluminium plate moves between, but does not quite touch,
the poles of a powerful permanent horse-shoe magnet, which
is attached rigidly to a support, which may be mounted on
the central pillar or independently. The effect is very similar
to the air damped balance. It has been stated that ferrous
materials can be weighed on this balance without appreciable
error.
GENERAL APPARATUS, REAGENTS AND OPERATIONS
In the following pages an account will be given of the
general -apparatus afld reagents of quantitative analysis.
Apparatus which is confined either to volumetric or gravimetric
;mal~sis will be described in the Sections devoted to these.
II, 10. General apparatus.-A. Burners. The ordinary
Bunsen burner is widely employed for the. attainment of
mod,rately high temperatures. It is important tl~at the air
regulator should be in working order. The maximum tem-
p~rature is attained by adjusting the regulator so as to admit
rather more air than is required to produce a non-luminous
flame; too much air gf'les a noisy flame, which is unsuitable.
An improvement in the design has been effected in burners
in which both the gas and air supplX .can be regulated. The
flow of gas is controlled at the base of the burner by means of
a screw which operates a needle valve; the ,Supply of air is
regulated by screwing the tube of the burner up or "down and
thus allowing more or less air to enter through the holes at (he
base . . The Pittshurgh universal bnmer* (sometimes termed
a Tirril bumer) and one of the Fletcher burnerst are of this
type. It is claimed that a temperature of l050-1100°C can
be produced in a covered platinum crucible with these burners.
With a M~ker burner a temperature of 1200°C is said to be
reathed. in a covered plqtinum crucible. The volume of air
passing thr9ugh a fully a-erated ordinary Bunsen burn..~r is
about 2·5 times the volume of the gas; for the c9mplete
combustion of coal gas about 6 volumes of air are necessary.
If a mixture of cpal gl!-s and air in the latter proportions ~ere
lighted at an ordinary Bunsen burner, the flame would" strike
back" and burn at t_Q.e bo~tom of the tube. In the Meker
burner (Fig. 2-7, a) the holes for the admission of air are
large enough to pass sufficient air for the complete combustion
of the gas, and the tube is narrowest near the base and widens'
out near the top, thus resulting in a more perfect mixing of
the gas with 'air; a ni<;:kel grid is fitted. into the top of the
burner in order to. prevent the flame striking back. The gas
* Manufactured by the Fisher Scientific Company, 711-723 Forbes Street,
Pittsburgh, Pa., U.S.A.
t Manufactured by Fletcher, Russell and Co. Ltd., Palatine Works, Warring-
ton, England.
Techniq1~e of Quantitative Analysis 219
\
burns in many small flames with the top of each inner reduc-
ing cone about 1 mm. above the top of the burner. The
•
-
(a) (b)
FIG. 2-7.
numerous small flames combine to give a very hot and highly

concentrated flame, which is oxidising in character except
below the tips of the tilly flames; the maxiplUm temperature
is attained just a little above the small flames, i.e.,'about 2 to 3
Jl?m. from the top of the burner. The 9.urner is used for the
ignition of precipitates .that require a high temperature for
conversion into a wei~bable form, and also for some fusions.
A blow-pipe flame or " blast lamp" is sometimes used for
these purposes, but in most cases equally satisfactory r~sults
are obtained with the Meker burner. An improved form of
the ordinary Miker burner (e.g., in the Fisher '''high tempera-
ture " burner Fig. 2-7, b) cont.ains an arrangement fur adjust-
ing both the ga~ and air supplies as in the Pittsburgh universal
burner already described.
The "Ama!" burner* attempts to combine the chief
features of th~ improved Bunsen burner and the Meker
burner. The flame can be turned down very low without
flashing back, and it also furnishes a very hot flame. This is
a very useful bumer, but is comparatively expensive.
B. Wash .bottles. A wash bottle is a flat-~ottomed flask
fitted up to deliver a fine stream of distilled water or' other
liquid ff;)r use ip. the transfer and washing of precipitates. A
convenient size is a 500 to 750 m!. flask of Pyrex or other
•
* Manufactured by Amal Ltd., Holford Works, Perry Bar, Birmingham,
England.
resistance glass; it should be fitted up as shown in Fig. 2-8.
A rubber bung is used, and the glass tubes above the bung
should be in the same straight line and
lie in the same plane. The jet should
deliver a fine stream of water; .a
suitable diameter of the orifice is I mm.
All glass tubing must be rounued in the'
Bunsen flame after cutting.
Thick string, thin sheet cork or other
insulating material, lield in place by
copper wire, should be wrapped round
the neck of the flask in order to protect
the hand when hot water is used. In
order to protect the mouth from scalding
FIG. 2-8.
by the back rush of' steam through
the mouth piece when the blowing is
stopped, it is convenient to use a three-holed rubber stopper
and a short piece of glass tubing open at both ends is inserted
in the third hole .• The thumb is kept over this tube whilst
the water is being blown 'Out, and is removed immediately
before th'e mouth,pressure is released!
All-glass wash bottles, fitted with ground-glass joints, can
be purchased. They should be used with organic solvents
that attack rubb_er.
C. Distilled water. Distilled water must be employed in all
analytical work. Various forms of still are in common use.
It is important that the vessel containing the ,.boiling water,
which may be of almost any material, should be so separated
from the condenser that there is little or no danger of spray
entering the latter. The condenser should be of pure tin or,
if this is not possible, ~e heavily plated with tin. Th~ dis-
tilled water thus prepared may contain small quantities of
dissolved ammonia and carbon dioxide. For water analysis
and other purposes requiring highly purified water, the dis-
tilled water should be re-distilled fro~ alkaline potassium
permanganate, and the head and tail fractions- rejected;
most of the dissolved ammonia is removed by this process.
The most co:q.venient method for the preparation' of highly
purified water (the so-called conductivity water) is to employ
a special type of still, such as that designed by the writer* ;
in this 'the steam is fractionally condensed and two grades of
conductivity water are simultaneously produced.
* Obtainable from Baird and Tatlock (London) Ltd., 16-17 Cross Street,
London. E.C.I. England.
Technique of Qu:antitative Analysis 221
D. Glassware. In order to a:'\'oid. the introduction of
impurities during analysis, resistance glass apparatus should
be employed. .For most purposes Pyrex glass (a boro-silicate
glass) is to b~ prefeqed; with strongly alkaline solutions,
however, the author finds Jena glass to be more resistant.
For elementary work, Chance's glass or Chance's" Hysil "
glass, which is che.aper, i;:; quite satisfactory. As a rule
glassware should not be heated with a naked flame; they
should be heated on a wire gauze, preferably with an asbestos
centre.
The most satisfactory beakers for general use are those
provided with a spout. The advantages of this form are:
(a) convenience of pouring, (b) the spout forms a convenient
place at which a stirring rod may protrude from a covered
beaker, and (c) the spout forms an outlet for steam <?r escaping
gas when the beaker is covered with an ordinary clock glass.
The size of a beaker must be selected with due regard to the
'{olume ofthe liquid which it is to contain. The most useful
sjzes are from 250 to 600 ml. For evaporations with high
boiling point liquids, e.g., sulphuric acid, the shallow squat
type of beaker is preferable.
COI).ical (or Erlenmey.er's) flasks of 200 to 500 ml. capacity
find many applications, for example., in titrations.
Funnels should enclose an angie of 60°. The most useful
sizes for quantitative analysis are those with diameters of
5'5,7 and I) cms. The stem should have an internal diameter
of about 4 mm. and should be no~ mqre than 15 crn. long.
For filling burettes and transferring solids to graduated
flasks, a short stem, wide-necked funnel is useful.
-E. Porcelain apparatus. Porcelain is generally employed
for operations in which hot liquids are to remain in contact
with the vessel for prolonged periods. It is usually considered
to be more resistant to solutibns, particularly alkaline solu-
tions, than glass, although this will depend primarily upon
the quality of the glaze. Shallow porcelain basins with lips
are employed for evaporations. Casseroles are lipped, flat-
bottomed porcelain dishes provided with handles; they are
more convenient to use than dishes.
Porcelain crucibles are very frequently utilised for igniting
precipitates and heating small quantities of solids because of
their cheapness a,nd their ability to withstand high tempera-
tures without appreciable change. Some reactions, such as
fusion with sodium carbonate or other alkaline substances,
and also evaporations with hydroflutnic acid canno,t be car-
ried out in porcelain crucibles owing to the resultant chemical
attack. A slight attack of the porcelain also takes place with
pyrosulphate fusions. •
F. Fused silica appMatus. Twq varieties of silica apparatus
are available commercially, the translucent and the trans-
parent grades. The former is much cheaper and can usually
be employed instead of the transparent variety. The ad~an
tages of silica ware are: (a) its great reslstance to heat shock
because of its very small coefficient of expansion, (b) it is not
attacked by acids at a' high tempe~ature except by hydro-
fluoric acid and- phosphoric acid and, (c) it is more resistant
to pyrosulphate fusions than is porcelain. The 'chief disad-
vantages of silica are: (a) it is attacked by alkaline solutions
and particularly by fused alkalis and carbonates, (b) it is more
brittle than ordinary glass, and (c) it requires a much longer
time for .heating and cooling 'than does, say" platinum
apparatus., /
G. Platinum appalatus. This is discussed fully in Section
II, 37. Mention may, however, be made here of the resistance
of platinum ware to fusions with sodium and/or potassium
carbonate, and to evaporations wjth hydrofluoric acid and
other single acids. A gre~t advantage is it~ high thermal
conductivity: a red hot platinum crucible placed in a desic-
cator to cool is ready for weighing after about 25 minutes.
H. Silver apparatulL The chief uses of silver crucibles and
dishes in the laboratory are in the evaporation of alkaline
solutions and for fusions with caustic alkalis; in the latter
case, the silver is slightly attacked. Silver melts at 960 0 and
care should therefore be taken when heated over a bare flame.
I. NiCkel ware. Crucibles and dishes of nickel are employed
for fusions with alkalis and witli sodiu'm peroxide. In the
peroxide fusion a'little nickel is introduced, but this is usually
not objecfronable. No metal entirefy withstands the action
01 fused sodium peroxide. Nickel oxidises in air, hence
nickel apparatus cannot be used for operations involving
weighing. •
J. Iron ware. Iron crucibles ~ay be substituted for those
of nickel in sodium peroxide fusions. They are not so
durable, but are much cheaper. ..
K. Bakelite apparatus. This synthetic resin is unaffected
by hydrofluoric acid at the ordinary temperature. Bakelite
Technique 'of Quantitative )1nalysis 223
funnels, measuring cylinders, etc., are available for use with
this acid.
L. Stirring rods. These are made from glass rod 3 to 5
mm. in diameter, cut into suitable lengths. .Both ends should
be rounded by heating in the Bunsen or blowpipe flame.
The length of the stirrihg rod should be suitable for the
size and the shape of the vessel for which it is employed, e.g.,
for use with a beaker provided with a spout, it should project
3 to 5 cm. beyond the lip wh~ in a resting position.
A stirring rod of convenient size should be provided with a
short piece of rubber tubing. fitted tightly over one end. This
is the so-called policeman; it is used for detaching particles
of a precipitate adhering to the side of a vessel which cannot
be removed by a stream of water from a wash bottle; it
should- not, as a rule, b~ employed for stirring nor should it be
allowed to remain in a solution.
M. Boiling rods. Boiling liquids and liquids in which a gas,
such as hydrogen sulphide, sulphur dioxide, etc., has to be
removed by boiling can be prevented from
superheating a~d "bumping" by 'the' use of a·
boiling rod {Fig. 2-9). T..h.is consists of a piece'Of
glass tubing closed at one end and sealed approxi-
mately 1 cm. from the other ena; the latter end
is immersed in the liquid. When the rod. is
removed, the liquid in the open end must be shaken
out and the rod rinsed with a jet of water from a
wash bottle. This device should not be used in
solutions which contain a precipitate.
N. Desiccators. An ordinary Scheibler-type of
desiccator (Fig. 2-10, A) is used for maintaining
a dry atmosphere in which to place objects that FIG. 2-9.
(A) (8)
FIG. 2-10.
might be aff~cted by moisture 9r carbon dioxide. The
drying agent (usually calcium chloride in' elementary
work) is placed in the lower compartment, and the upper
compartment is' fitted up for the accommodation of cru-
cibles, weighing bottles, etc. For small desiccators, a silica
triangle with wire ends is bent so that the ends are
perpendicular to the plane of the triangle, and-then the wire
ends spread out sufficiently so that they press against tHe
walls and floor of the comparfment and thus held firmly into
position. With large desiccators, a porcelain plate on feet
(Fig. 2-10, A), provided with apertures for crucibles, etc., is to
be preferred: it should' be. wedged into the sides, if necessary,
with cork or some other material. The ground edge of the
desiccator should be lightly coated with white vaseline or a
special grea~ in ord«:r to make it air tight; too mtich ~rease
may permit the lid to slide. •
When a hot object is placed in a desiccator, about 5 to 10
seconds should elapse for the air to become heated and ex-
panded before putting the cover in place. In removing the
object, the cbver should be slid open very gradually in order
to prevent any sudden inrush of air due to the partial vacuum
which exists 'owing to the cooling oHhe expanded gas content
of the desiccator and to desiccation and thus prevent the
precipitate being blown out of the crucible. '
A desiccator is, also employed for thorough drying of solids
for analysis and other purposes. A very convenient type is
portray~d in Fig, 2-10, B, a form ot vacuum" desiccator;
II
these are made of thick glass and should be tested by the

manufacturers for use with a high vacuum, Large surfaces
of the solid can be exposed, and the whole can be evacuated;
drying is thus much more rapid than in the ordinary Scheibler
desiccator.
The most commonly used desiccant is granular, fused
calcium chloride; the 3-8 mesh size is very suitable, but the
fused lumps are often employed on account of cheapness.
In point of fact the so-;cal1ed anhydrous calcium chloride (the
hydrated crystalline variety is useless for this purpose) is not
~ very efficient drying agent (it usually contains- about 1/4
molecule of water), but it is satisfactory for most purpose.s.
Phosphorus pentoxide is the most powerful common desic-
cant known; it should be used only when it is absolutely
necessary, and must be renewed as soon as a crust forms over
its surface. Concentrated sulphmic acid is better than cal-
cium chloride; it should be poured over pumice or broken
Technique of Quantita#ve Analysis 225
glass, and must be replaced as soon as it shows signs of
- darkening. Resublimed iodine must not be dried over
sulphuric acid. Silica gel, which has been dried at 130°C
and allowed to coo~ over phosphoric oxide, contains about 11
per cent.of water and is an extremely useful desiccating agent.
In recent years two ne\\;; desiccants have found wide applica-
tion, viz., magnesium perchlorate trihydrate (U dehydrite ")
and anhydrous magnesium perchlorate (" anhydrone ").
They are comparatively expensive for use in desicca:tors;
anhydrone is, however, extensively employed in the quantita-
tive absorption of water. Anhydrone rivals phosphorus
pentoxide in its power of desiccation; it can absorb about 50
per cent of its weight of water, and can be regenerated by
suitable means. The fpllowing Table, due largely to J. H.
Bower (1934), contains results which indicate the comparative
drying powers of a number'1>f desiccants.
TABLE XXVI. COMPARATIVE EFFICIENCY OF VARIOUS DRYING AGENTS.
Residual Residual
Material water per Material water per
litre of air lit~e of air
in mg. mmg.
CuSO.. (anhyd.) 2·8 Silica gel 0·03

CaCI, (gran. anhyd., KOH (sticks) 0·014
tech.) 1·25-1·5 AltO. 0·005
ZnCl. (sticks)
Ba(ClO')2 (anhyd.)
0·98
0·82
CaSO, (anhyd.)
CaO
. 0·005
0'003
NaOH (sticks) 0·80 Mg(ClO.). (anhyd.) 0·002
CaCI, (pure, anhyd.) 0·36 BaO 0·0007
Mg(CIO,J., 3H.o 0·03 PaOa 0·00003
It is evident from this Table that a hygroscopic material

such as ignited alumina or lime sllould not be cooled over
calcium chloride; either anhydrone or phosphoric oxide is
suitable.
Attention is directed to an all-aluminium desiccator (th~
Fisher" desicooler ") which is now marketed.* It is fitted
with a shelf containing three holes to accommodate three
sizes of crucible, and employs activated alumina as the desic-
cant (corrosion of the aluminium may occur'" with other
desiccants). The charge of about 60 g. of activated a~umina
may be reactivated when exhausted' by simply heating to
! 75°C. This desiccator should prove useful for general
students' use since it is unbreakable and inexpensive .
•
• Obtainable from the Fisher Scientific Co •• Pittsburgh. Pa .• U.S.A.
O. Test papers. With the development of more refined
methods of quantitative analysis, the use of litmus paper as a
rough indicator for determining the acidity and alkal~nity of
solutions has become obsolete. Paper strips coated with
indicators are now ,available commercially* in books or in
cork-stoppered glass vials. These iq,clude the following (the
pH ranges are given between the parentheses): thymol
blue (1·2 to 2'8); bromo-phenol blue (2'8 to 4'6); congo red
(3·0 to 5·0); methyl red (4·2 to 6·3) ; propyl retl (4·6 to 6·6)';
bromo-cresol purple (5·2 to 6·8); bromo-thymol blue (6'0 to
7 ·6); phenol red (6·8 to 8·4); cresol red (7·2 to 8·8); and
ortho-cresolphthalein (8'2 to 9·8).
The most outstanding advance, however, is the Fisher
" alkacid" paper,t which will .indicate the approximate
hydrogen ion concentration over the pH range 2 to 10 to
about ± 1 pH unit. Thus such terms as strongly acid, weakly
acid, strongly alkaline or weakly alk-aline assume a real
significanc~ With this test paper. When an "alkacid"
test paper is immersed in a solution, it assumes the following
colours: red at pH 2, orange at pH 4, yellow at pH 6, green at
pH 8, and blue at pH 10. Each vial Qj test papers contains a
5-colour chart printed'on the label, whichfacilit:ites matching.
P. Centrifuge. A sma11 electrically-driven centrifuge is a
useful apparatus for an analytical laboratory. It may be
employed for removing the mother liquor from recrystallised
salts, for coliecting difficultly filterable precipitates, and for
the washing of certain precipitates by decantation.
n, 11. Reagents.-The purest available reagents, both
liquid and solid, should be used for quantitative analysis.
Th~ analytical reagent quality is generally employed. Most
manufacturers now state the various percentages of impurities
on the label, and some give a guar~mtee that certain impurities
do not exceed certain definite limits.! It must be remem-
* These may be obtained, for example, from E. H. Sargent and Co., Chicago,
U.S.A., or from Hopkin and Williams Ltd,., London, E.C.2, England.
1 Supplied by the Fisher Scientific Co., Pittsburgh, Pa., U.S.A.
t See Analar Standards for Laboratory Chemicals, Second Edition, 1937,
published jointly by Hopkin and Williams Ltd., 16-17 Cross Street, London,
E.C.l, and by British Drug Houses Ltd., Graham Street, London, N.1, England.
See also the Analytical Reagents Catalogues of the Mallinckrodt Chemical
Works, of 72-74 Gold Street, New York, N.Y., U.S.A., of the J. T. Baker
Chemical Co. Phillipsl:iurg, N.J., U.S.A., of Baker and Adamson, 40 Rector
Street, New York, U.S.A., and of The Coleman and Bell Co., Norwood, Ohio,
U.S.A. These American firms, particularly the.first-named, have a very wide
range of analytical reagents and many conform with the specifications for
analytical reagents issued by the American Chemical Society (see Appendix
Section A,9). •
bered that the label is not an infallible criterion: for the follow,;
ing reasons:
(i) Some impurities may not have been- tested fQ.r by the
manufacturer.
(ii) In spite of the greatest care exercised in packing, foreign
materials, such as dust, small pieces of cork, paper, etc., may
be accidentally introduced.
(iii) The re~gent may have been contaminated after its
receipt from the manuf~cturers eithf!r by- the stopper having
been left open for some time with the consequent exposure
of t~e contents to the laboratory atmosph~re or by the inno-
cent return of an unused portiop. of the reagimt to the bottle.
(iv) In the case of a solid reagent, it may not be sufficiently
dry. This may be due eithe~ to insufficient drying by the
manufacturers or to leakage thmugh the stoppers during
storage, or to both of these causes.
• However, if the analytical reagents be purchased from a
manub.ctunng firm of repute, and instructions given (a) that
no bottle is to be 'opened for a longer time than is absolutely
necessary, a;nd (b) that no. reagent is to be returned to the
bottle after it has been removed, the likelihood of any errors
arising ftom some of the above possible causes is considerably
reduced. The exact .mode of drying, if necessary, will vary
with the reagent; details are given in the text in connexion
with their use. In- all cases of doubt as to the purity of the
reagents used, they should be tested by standard methods*
for the impurities that might cause errors in the determina-
tions. It must be mentioned that not all chemic~s employed
in quantitative analysis are available in the form ,of analytical
reagents; the purest commercially available products must
then be purified by known fnethods, some of wltich are given
below. •
For expressing concentrations of reagents', the molar sys-
tem is universally applicable, i.e., the number or fraction of
gram molecular ~eights present in 1 litre of solution. 'Con-
centrations may also be expressed in terms of normality if no
Gmbiguity is likely to arise (compare Section I, 23). All
reagent bottles should be kept scrupulously clean, particu-
larly round tne neck and mouth. The stopper should never
be laid on the bench, but sh?uld be held between the fingers or
placed upon a ,clean watch glass; it should be returned
* See, for example, J. Rosin, Reagent Chemicals and Standards, 1937 (Chap-
man and Hall).
228
.
Qua..ntitative Inorganic Analysis
immediately to its bottle as soon as the reagent has been
removed.
n, 12. Purification of substances.-Space does not permit
more than a short discussion of a few of the most important
methods of purification which are likely to be, employed by the
student.
A. Recrystallisation of solids. This is the commonest m'ethod
of purification; the usual solvent is water. A known
weight of the solid is dissolved in a volume of water sufficient
to give a saturated or nearly saturated solution at the bojling
point: a beaker, conical flask or porcelain· dish may be used.
The hot solution is filtered through a fluted filter paper placed
in a funnel with a short stem, and the filtrate collected in a
beaker: this process will remove insoluble material which
is usually present. .If the substance crystallises out in the
funnel, it should be filtered through a hot water funnel. In
the latter, the conical portion of the funnel is surrounded'
by a double-walled copper vessel containing Water, which
can be heated either by placing a ·Bunsen burner beneath a
hollow projection in direct communication with the copper
jacket or, better, electrically. The clear hot filtrate is
cooled rapidly by i.mmersion
in' a di$h of cold water or
in a mixture of ice and
water, according to the
solubility of the solid; the
solution is constantly stirred
in order to promote the
formation of small crystals,
which include less mother
liquor and are therefore more
pure than larger crystals.
The solid is then separated
from the mother liquor
by filtration. The various
funnels which may be
employed fof this purpose
are shown diagrammatically
in Fig. 2:-1l. (a) is the
ordinary porcelain Buchner
funnel; (b) i s the ] e n a
(d)
Buchner" slit sieve" glass (c)
funnel. In both cases, one FIG. 2-11.
technique of Quantitative A'nalysis 229
.or, better, two good quality filter papers are placed on
the plate; the Jena type is preferable since it -is trans-
parent and it is easy to see whether the funnel is perfectly
clean. (c) is a Jena Buchner funnel with a sihtered glass
plate; no filter paper is required; strongly alkaline and
acidic solutions can be readily filtered with this funneL
In all cases the funnel of appropriate size is fitted into a
filter flask (d), and the filtration conducted under the
diminished pressure provided by a filter pump or vacuum
line. The solid may be pressed down on the funnel with a
wide' glass stopper, sucked as dry as possible, and then
washed with sma)l portions of the original solvent to remove
the adhering mother liquor. The recrystallised solid is dried
upon clock glasses at or above the laborafory temperature
according to the nature of the material; care must of course
be taken to exclude dust. Various methods for .drying at
temperatures above that of the room are described below.
The dried solid is preserved in 'glass-stoppered bottles. It
should be noted that unless great care 'is taken there is a
danger of introducing a little filter paper with funnels (a) and
(b), and small particles of glass with funnel (c) when the solid
is removed.
Some solids are either too soluble, or the solubility does not
vary sufficiently with temperature, in a given solvent for
direct crystallisation to be practicable. In !]lany cases, the
solid can be precipitated from, say, a concentrated aqueous
solution p,y the addition of a liquid, miscible with water, in
which it is less soluble. Alcohol, in which most inorganic
compounds are almost insoluble, is generally used. Care
must be taken that the amount of alcohol or other solvent
added is not so larg_e that the impurities are also precipitated.
Potassium bicarbonate and tartar emetic (potassium anti-
monyl tartrate) may be purified by this method.
B. Sublimation. This process is employed to separate
yolatile substan.ces from non-volatile impurities. Iodine,
arsenious oxide and ammonium chloride can be purified in
this way. The substance is placed in a porcelain dish or
casserole; the latter is gently heated with 'a small flame and
the vapour condensed upon a cool surface, such as a large
inverted glass funnel containing a plug of glass wool at the
apex, or, preferably, a flask containing cold water.
Pure iodi,te is required in volumetric analysis, and details
for its purification will not be out of place here. Grind
230 Quantitative Jnor~anic Analysis
together 10 g. of iodine with 4 g. of potassium iodide (the latter
to retain any chlorine or bromine present as the non-volatile
potassium salts) and transfer the mixture to a casserole:
place a flask almost filled with cold waterI
on the casserole a in Fig. 2-12. Heat

very gently until sufficient iodine has sub-
limed on to the bottom of the flask, allow
to cool, and remove the flask with the
iodine adhering to it. Pass a rapid stream
of cold water through the flask; this
will cau~e the glass to contract sbmewhat
and the whole of the crust can then be
removed by scraping with a clean glass
FIG. 2-12. rod and is collected on a clock glass. Break
up the large pieces, and repeat the sub-
limation without the addition of potassium iodide. Remove
the second. ~blimate as before, and grind the iodine in a glass-
mortar. Dry in a desiccator containing calcium chloride;
no grease whatever should be exposed on the inside, since
iodine vapour attacks greas~ forming hydrogen iodide:
II, 13. Heating apparatus.-A. Steam -baths. These are
used for the slow evaporation of liquids, digestion of precipi-
tates, etc. The simplest form is a lipped beaker in which
water is boiled, the vessel being supported on the rim.
Generally, a cupper vessel, about half-filled with water and
heated by a Bunsen burner, is employe<;l. The upper portion
is fitted with a series of flat, sheet copper rings of- gradually
diminishing diameter so that vessels of various sizes can .be
accommodated. If the water bath is a large one, the upper
portion is fitted with a cover containing holes 6f various sizes,
and each aperture may be provided with its own series of
concentric rings or with an individual cover: vessels of various
dimensions can thus be heated at the same time. I t is
advantageous to fit a constant-level device to the water
bath; the danger of running dryCand consequent overheating
is thus eliminated. In some laboratories, the water batll.,
may be connected to a direct steam supply.
B. Steam ovens. A steam ove)l is usually made of copper
and, with the exception of the door, is double-cased through-
out. The casing is filled to about-one 'fourth of its height with
water and is heated by means of a Bunsen or other burner.
The water is kept boiling gently and the steam is condensed
by a suitable condenser and returned as water to the oven.
•
In many laboratories the heating of steam ovens is carried
out simultaneously with the distillation of water in the so-
called combined steam oven and still. The temperature
inside a steam oven usually does not exceed 90° to 95°C.
C. Electric ovens. The most convenient type is an elec-
trically heated, thermostatically controlled drying oven hav-
ing a temperature range from room temperature tp about
250°-300°C. These are comparatively inexpensive and
should form part of the normal equipment of every analytical
laboratory. The temperature can be controlled to within
± 1-2°C. They are used principally for drying precipitates
or solids at comparatively low controlled t_emperatures.
D. Muffle furnaces. An electrically heated furnace .of
muffle form should be available in every well-equipped labora-
tory. The maximum temperature should be about 1200°C.
If possible, a thermo-couple and indicating pyrometer should
be provided; otherwise the ammeter in the circuit should be
calibrated, and a chart constructed showing ammeter and
corresponding temperature readings. Gas heated muffles
may also be used, but are not so convenient in practice; the
maximum temperature is about 1000~C. "
E. Hot plates. The electrically-heated hot plate, prefer-
ably provided with three controls, "Low," "Medium"
and "High,"Iis almost indispensable for analytical laboratories.
The heating elements and the internal wiring should be
totally enclosed; this protects them from fumes and spilled
lrquids. The three-heat rectangular. type is particularly
suitable where a, number of evaporations are to be carried out
simultaneously. , •.
F. Air baths. -For drying solids and precipitates at tem-
peratures up to 250°C in which acid or other corrosive vapours
are ·evolved,.an electric oven should not be used. An air
bath may be constructed from a circular metal (copper, iron,
or nickel) vessel, lined round the sides and at the bottom with
two "thicknesses, 1/8" thick, of asbestos board; it is advisable
to soak the asbestos in water, before placing it in position.
J'he cover consists of an asbestos board, 1/4" thick, provided
with a hole for the insertion of a thermometer supported upon
a cork. A silica triangle, the legs of "Which are 'appropriately
bent, is inserted inside the bath for supporting a clock glass,
etc., The whole is heated by a Bunsen flame, which is
shielded from draughts; the temperat~re can be readily main-
tained' constant
- ' with 2 to 3°C. Recently. an air bath. of
232 Quantitative Inorganic {1 nalysis
similar construction, but with special J ena: glass sides (Fig.
2-13) has been introduced; this possesses the obvious advan-
tage of visibility inside the air bath.
- A form of air bath, which is very
useful for the speedy evaporation of
liquids (including concentrated sUlphuric
acid).at almost any temperature and also
for the dehydration of many solids, is
known as a radiator. I t is constructed
of sheet iron or nickel bent info the form
of a cone (Fig. 2-14, A) ; t"fte bottom B,
also - of nickel or iron, but best of
FIG. 2-13. platinum, is fixed to the cone by
bending the notched edges over one
another. A convenient size is 7 em. high, 7 cm. wide at the
top and 5 cm. at the bottom. A triangle of silica, porcelain
or platinum is suspended inside the radiator either by cutting
three narrow slits to a suitable
o
depth and placing the wire ends
of the triangle through these
(as in Fig. 2-14, q or the wire
ends may be bent over the
upper rim of the cone. * The
crucible or dish is supported (A) (8)
on the triangle. The radiator
is mounted on a tripod and
heated directly by a burner; ..
B
alternatively, it may be heated
on a hot plate. Evaporations
may thus be carried out speed-
ily without bumping. When
very rapid evaporation is re-
quired, a metal ring with an (C) . (D)
FIG. 2-14.
opening somewhat larger than
the mouth of the crucible, etc., is placed upon the top of the
radiator.
A'nother device of similar character is due to]. J. Moroney.
It consists of a metal cylinder. which is usually made of a east-'
aluminium alloy; the iriside bottom is provided with three
small projections upon which the beaker rests. There is
• The apparatus, known as the Hillebrand-Willard crucible radiator, is
commercially available. It is constructed from orle piece of monel metal, and
is marketed by the Central Scientific Cp., 1700 Irving Park Blvd., Chicago.
Ill., U.S.A.
Technique oj Quantitative Analysis 233
only a small clearance around the beaker and the'evaporator
{Fig. 2-14, D). The beaker is thus heated on the sides as well
as the bottom, and has three small hot spots on the bottom
at which points the bubbles form when boiling takes place.
The general superheating of the bottom is thus eliminated,
together with the attendant "bumping." The Moroney
evaporator is available commercially*; it eliminates the bump-
ing experienced in evaporating heavy solutions, or hydroxides·,
sulphuric acid, and solutions containing a precipitate.
IT, 14. Sampling of solids.-Sampling is the process of
extracting from a large quantity of material a small portion
which is truly representative of the composition of the whole
material, This process is of great importance sinc~, if it is
not satisfactorily done, the labour and time spent in making
a: careful analysis of the sample may be completely wasted.
If the material is more or less· homogeneous, sampling is
comparatively simple. If, however, the material is bulky and
heterogeneous, sampling must be carried out with great care,
and the method will vary somewhat with the nature of the
bulk solid. A detailed discussion of the methods of sampling
is outside the scope of fhis. book.f
The underlying principle for the sampling of 'material in
bulk, say, of a truck load of coal or iron ore, is to select k large
number of portions in a systematic manner from different
p~rts of the bulk and then to combine them. This large
sample-of the total weight iS-Crushed mechanically, if neces-
sary, and then.shovelled into a conical pile. Every shovelful
must fallllpon the apex of the cone and the shoveler must
w.alk around the cone as he shovels; this ensures a compara-
tively 'even distribution.. T.he top of the cone is then flat-
tened out ::tnd divided into quarters. Opposite qllarters of
the pile are then removed, mb~ed to form a smaller conical
pile, and again quartered. This process is repeated, further
crushing being carried" out if necessary, until a sample of
suitable weight (say, 200 to 300 g.) is obtained. "
If the quantity of material is of the order of 2 to 3 kilos or
less, intermixing may be accomplished by the method known
'" Obtainable from the Fisher Scientific Company, Pitt!burgh, U.S.A.
t Readers are referred to the various publications of the British Standards
Institution. e,g., the British Standards Specifications. No. 403 for 1930. which
contains a'report on the sampling of small fuel. and incidentally discusses
some general principles of sampling; also to the references in Hillebrand and
Lundell. Applied Inorganic Analysis, 1929, p. 47, and-in W. W. Scott-No H.
Furman. Standard Methods of Chemical Analysis, Vol. II. p. 5, 1301 (5th
edition, 1939).
lilt
234 Quantitative Inorganic Analysis.
as "tabling." The finely divided material (s spread on the
centre of a large sheet of oil-cloth. or similar material: Each
corner is pulled in succession over its diagonal partner, the
lifting being reduced to a minimum; the particles' are thus
caused to roll over and over on themselves, and the lower
portions are constantly brought to the top of the mass and
thorough intermixing ensues. The sample may then be
rolled to the centre of the cloth, spread out, and quartered as
before. The process is repeated until a sufficiently small
sample is obtained. The final sample for the laboratory,
which is usually between 25 and 200 grams in weight, is
placed in an air-tight bottle.
The sampling of metals and alloys is best effected'by drilling
holes, preferably right through the central portion, of a repre-
sentative test piece; all the material from the hole must be
analysed as the fine dust II}ay differ in composition from th~
more ductile dtillings. .
Before analysis the representative sample is usually dried
at 105-110°C, or at some higher specified temperature if
necessary, to constant weight. The results of the analysis'
are then reported on the "dry" basis, yiz., on a material
dried at a. specified temperature. The loss in weight OIL
drying may ,be determined, and the results may be reported,
if d~sired, on the original " moist" basis; these figures will
only possess real significance if the material is not appreciably
hygroscopic and no chemical changes, other than the loss of
water, take place on drying.
In a course of systematic quantitative analysis, such as
that with which we are chiefly concerned in the present book,
the unknowns supplied for analysis are usually portions 'of
carefully analysed samples which have been finely ground
until uniform.
IT, 15. Crushing and grinding.-If the material is hard (e.g.,
a sample of rock), it is first broken into small
pieces on a hard steel plate with a hardened
hammer. The loss of fragments is prevented
by c{)vering the plate with a steel ring. or
in some other -manner. The small lumps
may be broken in a "percussion" mortar
(F.ig. 2-15). The mortar and pestle are
constructed entirely of hard steel. One or
two small pieces' are placed in the mortar,
and the pestle inserted into position; FIG. 2-15.
the iatter is struck lightly witl). a hammer until the pieces
have been reduced to a -coarse powder. The whole of the
hard substance may be treated in this manner. The coarse
powder is then ground in an agate mortar in small quantities
at a time .
.fi, 16. Solution of the sample.-Most of the synthetic
samples employed in the teaching of elementary quantitative
analy~s are soluble in water or in dilute acids. Naturally
occurring minerals and ores, various alloys, etc., must usually
be treated with various reagents in order to bring them into
solution. In such cases, the qualitative analysis will have
revealed the best procedure to adopt. Each case must be
considered on its merits; no attempt at generalisation will
therefore be qlade. We can, however, discuss the experi-
mental technique of the simple process of solution of a
substance in water or in aci~s, and also of insoluble
substances.
The sample is weighed out into a. beaker, and the beaker
immediately covered with a clock glass of suitable size (its
diameter sho,!!ld not be more than about I cm. larger than
that of the beaker) with its convex side facing downwards.
The bei:\.ker should have a spout in order to provide an outlet
for the escape of steam or gas. The solvent is then added by
pouring it carefully down a glass rod, the lower end of which
rests against the wall of the bea~er; the clock glass is dis-
placed somewhat during this process. If a gas is evolved
during the addition of the solvent (e.g., acids with carbona"tes,
metals, alloys, etc.), the beaker must be kept covered as far
as' possible during the addition. The reagent is then best
added by means of a pipette or by means of a funnel with a
bent stem inshted beneath the clock glass at the spout of the
beaker ;, loss by spirting or spray is thus prevented. When
the evolution of gas has ceased and the substance has com-
pletely dissolved, the. under side of the clock glass is well
rinsed with a stream of water from a wash bottle, care being
taken that the washings fall ~n to the side of the beaker and
not directly into the solution. If warming is necessary, it is
usually best to carry out the entire operation of solution'in a
conical (Erlenmeyer) flask with a small funn'1l.. in the mouth
(Fig. 2-HI, a); loss ofliquicfby spirting is thus prevented and
the escape of gas is not hindered,
It may ofte,n be necessary to reduce the volume of a solu-
tion, or sometimes to evaporate it completely to dryness.
Wide and shallow vessels are most suitable. since a large
surface is thus exposed and evaporation is thereby acceler-
ated. We may employ shallow beakers of resistance glass
~
y
(b) ... (c)
(a)
FIG. 2-16.
(e.g .• Chance's" Bako " beakers). Pyrex evaporating dishes.

porcelain basins or casseroles. silica or platinum basins.; the
material selected will depend upon the extent of attack of the
hot liq.uid upon it and uporrtbe constituents being determined
in the subsequent analysis. Evaporations should be carried
out on the steam bath or upon a low-temperature hot plate;
slow evaporation is preferable to vigorous boiling, since the
latter may lead to some mechanical loss in spite of the pre-
cautions to be mentioned below. During evaporations, the
vessel must be covered by a clock glass of slightly larger
diameter than the vessel, and supporfed either on a large all-
glass triangle or upon ·three small U-rods of glass hanging on
the rim of the container. Recently, a device known as
Fisher's "speedy-vap" beaker cover (Fig. 2-16, b)* has been
introduced; this is a great improvement on the ordinary
watch. or clock glass and is of great value for simple evapora-
tions. I t consists of a clock glass of resistance glass provided
with three fused-on glass ribs. Needless to say, at the end
of the evaporation the sides of the vessel, the lower side of the
clock glass (simple or modified forms), and the triangle and
glass hooks (if employed) should be rinsed with distilled water
into the vessel.
For evaporation at the boiling point either a conical flask
with a short funnel in the mouth or a round-bottomed flask
inclined at an angle of about 45° (Fig. 2-16, c) maybe employed;
in the latter the drops of liquid, etc., thrown up by the ebulli-
tion or by effervescence will be retained by striking the inside
of the flask, wltlst gas and vapour will escape freely: .
Substances which are insoluble or only partially soluble in
acids are brought into solution by fusion with the appropriate
"'Marketed by the Fisher Scientific Co., of Pittsburgh, U.S.A.
reagent. The most commonly used fusion t reagents, or
fluxes as they are called, are anhydrous sodium carbonate
either alone or, less frequently, mixed with potassium nitrate
or sodium peroxide, potassium or, sodium pyrosulphate, sodium
peroxide, sodiu~ or potassium hydroxide. The flux employed
will dep:md upon the nature of the insoluble substance. The
vessel in which fusion is ~ffected must be carefully chosen;
platinum crucibles are employed for sodium carbonate and
potassium pyrosulphate; nickel or silver crucibles for sodium
or potassium hydroxide; nickel, silver or iron 'crucibles
for sodium carbonate and / or sodium peroxide; nickel
crucibles for sodium carbonate. and potassiuIl'/- nitrate
(platinum is slightly attacked). To carry out the fusion,
a layer of the flux is placed at the bottom of the
crucible, and then the intimate mixture of the flux and the
finely-divided substance added; the crucible should be about
half-full, and should, generally, be kept covered during the
whole process. The crucible is very gradually heated at first,
and the temperature slowly raised to the required tempera-
ture. The final temperature should not be higher than is
actually necessary; any possible further attac~ of the flux
upon the crucible is thus avoided. When the fusion, which
usually takes 30 to 60 minutes, has been completed, the
crucible is grasped by means of the crucible tongs (see Section
U. 37) and gently rotated and tilted so that the molten material
distributes itself around the walls of the container and solidi-
fies there as a thin layer. This procedure greatly facilitates
the subsequent detachment and solution of the fused mass.
When cold, the crucible is placed in a casserole, porcelain
dish, platinum basin or Pyrex beaker (according to Hie nature
of the flux) and covered with water. Acid is added, if neces-
sary, and the vessel is covered with a clock glas(> (preferably
of the Fisher" speedy-vap " type), the "temperature raised to
95°-100°C, and maintained until solution is achieved.
Full experimental details for effecting fusions with sodium
carbonate and pota.ssium pyrosulphate are given in Sections
IV, 70B, and IV, 95, and in IV, 94G respectively.
TECHNlQUE OF VOLUME'l'RIC ANALVstS'
n.l'. Unit of volume.-The fundamental unit employed in

measuring the volume of a liquid is the litre. The litre is
defined as the volume qccuP.ied by one kilogram of water at the
temperature of its maximum density' (4°C) and subjected to
normal atmospheric pressure. It should be' rtoted that the
quantitY,of water which occupies one litf'e at 4°C has a mass of
one kilogram, a{.ld not an apparent weight in air of one.kilo-
gram. The apparent weight in air of one litre of water at
4°C will be less than one kilogram by the buoyancy effect of
the air (compare Section U"7); if the weighing is.made at the
ordinary temperature, the apparent weight will be still less
because of the lower density 9f water at temperatures above
4°C .. Likewise the volume of water will depend, but to a
lesser degree, upon the pres'sure to which it is subjected.,
These factors must 'be allowed for when determining the
volume of a piece of apparatus by weighing tIle water that it
contains or delivers.
The millilitre (ml.) is the thousandth.part of a litre., This is
the most convenient ulJ.it for expressing the capacity of
volumetric glassware.
The cubic centimetre (c.c., c.cm., cm. 3) is the volume
occupied by a cube each of whose edges is one,centimetre in
length. The most recent determinations have shown that
the relationship between the millilitre and the cubic centi-
metre is.:
1000 ml. = 1000·028 C.c.
In this book the unit of volume will be taken as the milli-
litre in accordance with the recommendation of the National
Physical Laboratory 'in their pamphlets of September 1924
and April 1934, and now almost, universally adopted. It is
true that the difference between the ml. and c.c. is only 28
parts per million and t~erefore 'well within the limits of error
•of most volumetric determinations; for this reason many still
use the terms c.c. and ml. synonymously. It must be. re-
membered, however, that the ultimate unit of volume is the
litre, which is based upon a determination of mass; it is
therefore more logical to adh~re to the millilitre for expressing.
capacities. *
* The recent discovery of deuterium oxide DP (" heavy water "). which
has a specific gravity at 20°C of 1·1059 and a temperature of maximum
238
Technique oj Quantitative Artalysis 239
II, 18. Standard temperature.-Prior to 1934, the standard
temperature in Great Britain was 15°C, but since the issue in
1934 of the" Report on Standard Temperature of Vo!umetric
Glassware" (No. 554) by the British Standards Institution a
st,!ndard temperature of 20°C has been almost universally
adopted. The latter standard temperature is employed in
America and on the continent of Europe. This temperature
most nearly represents the average temperature of the labora-
tory, or at least a temperature which is readily attainaDle.
If we take the coefficient of cubical expansion of glass as
0·000026 ml. per ml, per degree Centigrade, the following
Table gives the correction to be added when the sign is +,
or subtracted when the sign is -, to or from the capacity of a
1000 ml. flask correct at 20°C in order to obtain the capacity
at other temperatures.
TABLE XXVII. TEMPERATURE CORRECTION FOR 1000 ML. GLASS

FLASK FOR THE EXPANSION OF GLASS (STANDARD TEMPERATURE,
. 20°C). •
. .
Degrees C. I fio 10° 15° 20° 25° 30°
Ml. - 0·39 - 0·26 - 0·13 0·00 + O'l~ + 0·26
In the actual use of volumetric glassware for the measure-

ments of 'liquids, the expansion of the liquid must also be
taken into consideration if temperature corrections are to be
l1lade. Table XXVIII gives the correction~ to be added or
subtracted in order to obtain the volume occupied at 200(
by a volume of water which at the tabulated temperature is
contained in an accurate 1000 ml. flask having a standard
temperature of 20°C. It will be seen that the allowam. ! for
the expansion of water is considerably greater than that for
the expansion of the glass.
"
n, 19. Volumetric apparatus.-The most commonly used
apparatus in volumefric analysis a.re measuring flasks,
density of 1l·6°C, introduces a complicating factor. The normal propprtion
of deuterium oxide in ordinary water is about I part in 5,000; this will resu'lt
in a difference of about 22 parts per million in density fron' that of pure pro-
trium oiide H 20.' Fortunately, the proportion of deuterium oxide does not
vary appreciably with all ordinary samples of water, but it is evident that a
precise definition of volume must specify the proportion of this constituent.
_ 240 Quantitative Inorgani"C Analysis
TABLE XXVIII. TEMPERATURE CORRECTIONS FOR VOLUMES OF
WATER MEASURED IN A 1000 ML. GLASS FLASK (STANDARD TEM-
PERATURE, 20° C).
Degrees C. 5° 10° 15° 20° 25° 30°

Correction
(m!.) + 1·37. + 1·24 + 0·77 0·00 .-1·03 - 2·30
burettes, and pipettes. Graduated cylinders and weight

pipettes ate less widely employed. Each of these will be
described in turn, together with the methods of use, calibra-
tion, and tolerances allowed by the National Physical
Laboratory.
Cleaning of glass apparatus. Before describing volumetric
apparatus in detail, reference must be made to the important
fact that all such glassware must be perfectlyJclean and free
from grease, otherwise the results will be unreliable. The
test fpr cleanliness of glass apparatus is that on bying filled
with distilled water and the water withdrawn, only an un-
broken film of water remains. If the water collects in drops,
the vessel is dirty and must be cleaned. Various methods are
available for cleaning. The method which is most frequently
used consists· in filling the apparatus with" cleaning mixture,"
a solution of powdered .sodium or potassium dichromate in
cQncentrated sulphuric acid, and allowing it to .stand for
several hours, but preferably overnight; the acid is then
poured off, the apparatus thoroughly rinsed with distilled
water, and allowed to drain until dry. A more expeditious
method of drying is to rinse the apparatus with· pure acetone
after draining off the water and then to blow or draw air
through it: under no circumstances should the apparatus
be heated. A more efficient cleaning liquid is a mixture of
concentrated sulphuric acid and fuming nitric acid; this may
be used i.f the vessel is very greasy and dirty. Grease may
also be removed by filling the apparatus with warm soap
solution, leaving for 15 minutes, rinsing with water, followed
by concentrated/hydr'ochloric acid, and finally with distilled
water.
n, 20. Measuring flasks. * A measuring flask is a flat~
bottomed, pear-shaped or npherical vessel with a long narro,W
• These are sotnetimcs knQWll a:;; volutnetric flasks.
. .
Technique of Quantitative Analysi~ 241
neck (Fig. 2-17). A thin line etched around the neck indicates
the volume that it holds at a certain d,efinite temperature,
usually 20°C (both the capacity and the temperature are
clearly marked on the flask); the flask is then said to be
graduated to contain. Flasks with one mark are
always taken to contain the volume specified.
A flask may also be marked to deliver a. specified
volume of liquid under certain definite conditions;
these are, however, not suitable for exact work.
Vessels standardised at the National Physical
Laboratory are marked with a letter D when
they are meant to deliver; if they are intended
both to contain and to deliver, the letter :P is
placed above the upper mark and the letter C
below the lower mark. .. .
The mark extends completely around the FIG. 2-17.
neck ire order to avoid errors due to parallax
when making the nnal adjustment; the lower edge of the
meniscus should be tangential to the graduation mark, and
both the front and the back of the mark should be seen as a
single line. The neck is made narrow so that a small change
in volume will have a large effect upo~ the height of the
meniscu.s: th~ error in adjustment of -the meniscus is
accordingly small. Volumetric flasks should be fitted with.
ground-glass stoppers, -and both the stopper and the flask
should be numbered.
Measuring flasks are ordinarily made of 50, 100, 250, 500,
1000, and 2000 ml. capacity. They are employed in making
up standard solutions to a given volume; tbey can also be
used for obtaining, with the aid of pipettes, aliquot portion~
of the substance to be analysed. .
Calibration. A large balance, which will ca:rry about a kilo-

gram in each pan and which has a sensitIvity when fully
loaded of about 5 mg., is required. The flask is first thor-
oughly cleaned and dried; it is then weighed on the large
balance. The flask is then filled with distilled water, which
has been standing in the balance room jar at least an hour
before use, to a few millimetres above the mark: care is
taken to avoid wetting the neck above the water surface.
The lowest point of the water meniscus is then adjusted to the
graduation mark by withdrawing small quantities of water
with 'a capillary tube or with a narrow tube drawn out to a
242 QuanMtative I nOTGan£c A nalys£s
,
jet at the lower end. The flask is then weighed,* and the
temperature of the. water noted. The true volume of the
water filling the flask to the graduation mark can be calcu-
lated with the aid of the following Table. The values in the
Table have been obtained by making allowance for (a) the
difference in volume of the glass vessel at the calibration
temperature and at 20°C, '(b) the density of water at the tem-
perature of the calibration, <i-nd (c) the effect of buoyancy of
the air upon the water and the brass weights. The figures
apply to an atmospheric pressure of 760,mm. of mercury and
a relative humidity of the air of 50 per, c~t; the usual
deviation from these figures will affect the buoyancy correc-
tion (compare S~ction n, 7) only slightly and can be neglected
for most purposes.
TABLE XXIX. WEIGHT OF WATER (BRASS WEIGHTS I.tI AIR) TO GIVE ONE
LITRE AT 20°C. I
Volume of Volume of
Temp. Weight 1 g. of water • Temp. Weight 1 g. of water
(0C) (g.) (mI.) (0C) (g.) (mi.)
100 998·39 1·0016 23° 996·60 1'0034
llO 998·32 1'0017 24° 996·38 ,1'0036
12° 998·23 1'0018 25 0 996-17 1'0038.
13 0 . 998·14 ' 1'0018• 26° 995'93' ['0041
140 998·04 1'0019 27° 995·69 1·0043
15° 997·93 1·0021 28° 995·44 1·0046
16° 997·80 1·0022 29° 995·18 1'0048
17° 997·66 1·0023 30· 994·91 1·0051
18° 997·51 1·0025 31° 994·64 1·0054
19° 997·35 1·0026 32° 994·35 1'0057
20° 997-18 1'0028 33° 994·06 1'0060
21' 997·00 1·0030 34° 993·75 1'0063
22° 996·80 1'0032 35° 993·45 1·0066
Note. For the calibration af flasks of capacity other than 1 litre, the
corresponding multiple or sub-multiple of the above values is taken.
If the error, calculated with the aid of Table XXIX, falls'

within the limits of tolerance of " Class A" apparatus (see
• If a precision type of balance is not available, or the eqqaltty of the length
of the balance arms is suspect, it is best to weigh by the method of substitution
(Section n. 60). Counterpoise the empty flask either by adding leat! shot into
a beaker plac~ on the right-hand pan and/or another set of weights: observe
the rest point. Remove the flask a~d replace it by weights sufficient to bring
the balance to the same rest point. Record the weigllt, and remove both
weights and tare. ""Weigh the flask filled with distilled water to the mark by
the method of substitution as before. The difference in weights on the left-
hand pan will ~ive the weight of the water,
following paragraph). the flask may be taken as accurate for
all ordinary quantitative work. If~ however. the error is
large. the simplest procedure is either to affix a small paraffined
label to the flask bearing the true capacity and the date of
calibration. or to calibrate another flask. TheJ process of
re-graduating the fl;tsk is best left to the manufacturers.
Standardised flasks. Measuring flasks which. have been
verified at the National Physical Laborato~y can be pur-
chased. The N.P.L. recognise two types of volumetric
glassware: those which have passed "Class A tests II
(marked liB.A.) and those which have passed "Class B"

tests (marked m.B.). The former are designed for vessels
intended to possess the high accuracy normally required for
scientific use; the latter are intended for orainary .routine
or commercial analysis. Apparatus which has pas~ed the
" Class A tests can also be supplied with a certificate giving
II
the exact volume; these may be employed in work demanding

the highest accuracy. Full details concerning the construc-
tion and design of volumetric glassware are given in the
pamphlet issued by the N.P.L. in April 1934.* Only the
tolerances or permissible errors for flasks" to contain will be II
given here.
1_
II
Capacity, mi. -
25 ~o ~oo 250 ~oo
~--
1000
~'2
"Class A" 0·03 0'04 0·06 0·1 0·15 0·2 0-4
Tolerance. ± mi. }"Class B n 0.04 0:06' 0·1 I 0·3 0·4 0·6
II, 21. Pipettes.-Pipettes are of two kinds: (i) those

which have one mark and deliver a small. constant volume of
liquid under certain specified 'conditions (transfer pipettes);
(ii) those in which the stems are graduated and are employed
to deliver various small volumes· at discretion (graduated.
or measuring pipettes). They are sho}Vn in Fig 2.18 as A and
C; type B is a useful form which is employed for measuring
01,lt corrosive liquids, the bulb preventing the entrance of
liquid into the. mouth. Graduated pipettes are useful for
measuring out volumes of liquids approximately; they are
not emplgyed in exact work.'
• The U.S. Bureau of Standards issue a similar pamphlet (Circular No.9
for 1916). Only one type of apparatus, corresponding roughly to the N.P.L.
'~Cl~ A •• is. however. allowed.
244 Quantitative Inorganic A nalys~s
Transfer pipettes are constructed with capacities of I, 2,
5, 10, 20, 25, and 50 mI.·; those of 10, 25, and 50 ml. capacity
. are most frequently employed.
In using such pipettes, they
are first rinsed with the liquid,
then-filled by suction to about
1-2 cm. above the mark, and the
upper end of the pipette is closed
with the tip of the dry index
25 finger; any adhering liquid is
ml
20
wiped from the outside of the
C. lower stem. The liquid is allowed
to run out slowly by slightly
relaxing the pressure of the
finger and by ~arefully rotating
the pipette until the bottom of
the meniscus just reaches the
graduation _mark; the pipette
must be held vertically so that
the mark is at the same level as
0) (8) (C) the eye. Any drops adhering to the
FIG. 2-18. tipareremovedbystrokingagainst
a glass surface. The liquid· is
then allowed to run into the receiving vessel, the tip of the
pipette touching the wall of the vessel. When the continuous
discharge has' ceased, the jet is helG in contact with. the
side of the vessel.Jor 15 seconds (draining time). At the ·end
of the draining tim.e, the receiving vessel is removed from
cont~ct with the jet of the pipette, thus removing any drop
adhering to the outside of the pipette. The liquid remaining
in the jet of the pipette must not be removed either by blowing
or by other means; blowing .may introduce grease from the
mouth. The above standard procedure for filling and empty-
ing a pipette is that recommended by the National Physical
·Laboratory, and will be employed in its calibration.
A pipette will not deliver constant volumes of liquid if
discharged too rapidly. The orifice must be of such size that
the time of outflow is about 20 seconds for a 10 m!. pipette,
30 seconds for a 25 ml. pipette, ~nd 35 seconds for a 50 ml.
pipette.
Calibration. The pipette is cleaned by filling it with di-
chromate-sulphuric acid cleaning mixture and allowing it to
stand, preferably overnight. J!illing is effected by attaching
a short piece of rubber tubing to the upper end and either
immersing the pipette almost comp~etely in a tall cylinder
containing the cleaning mixture, or by direct suction, a safety
bottle (Fig. 2-19 ; a glass-stoppered gas wash
bottle serves equally well) being inserted
between the pipette and the mouth. The
pipette is''allowed to drain, thoroughly washed
with tap water, and finally with distilled water.
The pipette is then filled with distilled water,
which has been standing in the balance room
FIG. 2-19.
for at least an hour, to a short distance
above the mark. Water is run out until the meniscus is
exactly on the mark, and the outflow is then stopped. The
drop adhering to the jet is removed- by bringing the surface
of some water contained in a beaker in contact with the jet,
and then removing it without jerking. The pipette is then
allowed to d~charg.e into a clean, weighed stoppered flask
(or a large weighing bottle) and held so that the jet of the
pipette is in contact with the. side of the vessel (it will be
necessary to slightly incline eith~r the pipette or the vessel).
The pipette is allowed to drain for 15 seconds after the out~
flow has ceased, the jet still being in contact with tne side of
the vessel. At the end of the draining time the receiving
vessel is removed from contact 'with the tip of the pipette,
thus removing any drop adhering to the outside of the pipette
and ensuring that the drop remaining in the end is always of
the same size. To determine the instant at which the outflow
ceases, the motion of the water surface down the delivery
tube of the pipette is observed, and the delivery time is
considered to be complete when the meniscus comes to rest
slightly above the end of the delivery tub·e. The draining
time of 15 seconds is counted from this moment. The receiv~
ing vessel'is weighed, and the temperature of the wat~r noted.
The capacity of the pipette is then calculated with the aid of
Table XXIX. At least two determination should be made.
If the error is within" Class A" tolerance (see below), the
pipette may be acce:eted at its nominal value. If the error
is larger, a small paraffined label may be 'attached to the
pipette indicating the error (this, is the simpler procedure),
or a new graduation may be made as follows. A thin strip
of gummed paper is attached to the pipette and a mark made
where it is thought that the graduation should be;· a rough
guide is provided by noting how far. the wate~ falls in the
narrow. tube .when one drop (roughly 0·05 ml.) 1S run out of
246 Quantitative I norganie Analysis"
the pipette. The new graduation is then tested as before.
If this is not quite correct, the exact position can now be
readily determined. The fillal mark can be made permanent
by etching a ring into the glass with hydrofluoric acid.
Standardised pipettes. Pipettes which have been tested at
the-National Physical Laboratory may be' purcha,.c;ed. The
times of outflow and the tolerances for the .. Class A " and
" Class B " tests* are tabulated below.
Capacity, ml. 2 10 20 30 .50 1100

Time of outflow,
sees., " Class A " 7-15 18-25 20-35 20-35 25-40 30-50
- -- -- -- - --. - -
Tolerance, ± mI.,
"Class A .. 0·01 0·02 0·02 0·03 0·04 0·06
----------
Time oj" outflow,
sees., .. Class B " 7-20 15-40 20-50 20-50 25-60 30-80
----------
Tolerance, ± mI.,
.. Class B" 0·02 0·04 0·05 0·06 0-08 0·12
,
Only those of'" Class A" are suitable as standards for
precise quantitative work. "Class ~ " pipettes are marked
with reference to delivery (D), time of outflow, and of drainage
• (15 seconds). A typical inscription for a 50 mI. pipette would
be:
50ml.
D.2000
(30 + 15) Sec.
II, 22. Burettes.-Burettes are long cylindrical tubes' of

uniform bore throughout the graduated length; they are
closed at the: bottom by means of a glass stopcock, or with a
short piece of india-liUbber tubing carfying a glass jet (Fig.
2-20, A and B). The rubber tubing is closed either by means
of a pinch-cock or by means of a glass bead; by firmly squeez-
ing the rubber surrounding the bead between the thumb and
the fore-finger a narrow channel is opened, through which the
liquid escapes, the rate of flow being readily controlled. Ty.pe
•
• See Section D, 20 for explanation of " Class A " and" Class $."
•
C (Fig. 2-20) burettes are generally employed for titrating ii"ot
solutions; the heat will not reach the bulk of the solution in
the burette and thus will not affect its volume. Burettes
are usually of 50 ml. capacity graduated to tenths of a ml. :
<l; eo
- :-
- :.
: :
:
·
·
-
·
:
·:
::_ ·
·
~
·
c:-
-
(D)
-
·
--
-·
(8)
FIG. 2-20.
capacities of 10 rul. and 25 ml. are also available. Burettes

are employed. to deliver variable volumes of liquid.
The mode of use is as follows. The clean, empty burette is
.fixed Into a suitable stand and clamped vertically. * About
10 m!. of the solution to be employed is poured into it so as to
* Of the many stands - on the market the "Technico" burette holder
(Registered Design, No. 830030), illustrated in Fig. 2-20 (burettes A and Bare
held in position by this double clamp), is very convenient a~d also inexpensive.
No parts of the holder obscure graduation lines pr numbers on the burette, and
the burette is held firmly in a truly vertical position. The movement of the
burette in order to bring the meniscus to eye level is but the work of a moment.
A single burette holder, which can be fixed on aretort stand (Fig. 2-2 I, D), is
also marketed. These burette \ holders may be purchased fro m the sale
manufacturers, Messrs. A. Gallenkamp and Co. Ltd., 17-29 Sun Street, Fins-
bury Square, London, E.C.2, England. •
wet the whole internal surface, and is then run completely
out; this rinsing is repeated. The burette is then filled to a
little above the highest graduation; if a short funnel is used,
it must be removed -immediately after the filling. The liquid
is then allowed to flow out until the lowest· point of the menis-
cus just touches this graduation; the jet must of course be
completely full of liquid. To read the position of the menis-
cus, the eye must be at the same level as the meniscus in ord~r
to avoid errors due to parallax. In the¥best type of burette,
the graduations are carried completely round the tube for
each m!. and half-way round for the other graduation marks:
parallax 'is thus easily avoided. To aid the eye in readi}1g t}le
position of the meniscus a piece of white paper or cardboard,
the lower half of which -is blackened either by -painting with
dull- black paint or by pasting a piece of dull black paper
. upon it, is employed. When
this is placed so that the sqarp
diyiding line is 1 to 2 mm. below
the meniscus, the bottom of
the meniscus appears to be
darkened and is sharply out-
lined against the white back-
ground (Fig. 2-21)* ; tlie level
of the liquid can then be accur-
ately read. For all ordinary
FIG. 2-21. purposes readings are made to
0·05 ml.; for precision work
readings to 0·01-0·2 ml. (by estimation with the aid of a lens)
should be 'made.
The so-called Schellbach burette, which contains a white
enamelled band, with central blue ribbon, fused vertically at
the back of the burette tube, was originally intended to im-
prove the ease of reading the level of the liquid and also to
eliminate parallax. Its use in· practice is not, however, very
satisfactory (such burettes are admitted to only class" B "
test.s at the N,P.L.), and hence cannot be recorpmended.
AS a general rule, it is advisable to use the zero mark each
time, It has been established that practically no after-
drainage takes place in a burette if. the rate of outflow does
'not exceed a certain speed. If the speed is exceeded, owing
to the jet being too large, 'then drainage takes place for
• A celluloid black and white disc, which Can be slipped over the burette is
a.vailable commercially (e.g. " Fisher" burette meniscus holder, "Baird and
Tatlock" burette reader, etc.).
several minutes. The limits for the time of outflow with the
tap fully open and the jet not in contact with the receiving
vessel are given in connexion with the N .P.L. "Class AI/
tests' in a subsequent paragraph. A burette should be
selected which conforms to these limits. The drop adhering
to the jet after the liquid has been discharged is removed by
bringing the side of the receiving vessel into contact with the
jet. When not in use, burettes should be well rinsed with
distilled water and covered with a short inverted test-tube
to keep out dust, or they may be inverted in the burette
stand, jet uppermost. •
Burettes fitted with rubber and a pinch cock are used for
alkaline soh~tions and as cheap substitutes for glass stop-
cock burettes; they cannot "Qe employed for such solutions
as potassium permanganate and iodine, which attack rubber.
The use of any form of burette other than that fitted with a
glass stop-cock is not recommended for the following reasons :
(i) some of the liquid may attack the rubber, and (ii) it is
impossible to' see whether the air has been completely dis-
placed in the rubber tube. Alkaline solutions, including
those of caustic alkalis, may be ~sed with glass taps, but the
burettes must be emptied and washed immediately after use,
otherwise the taps are liable to stick. Where the use of
alkaline solutions in a burette is a routine process,1t is best
to utilise a nickel or silver stop-cock.
Lubricants (or greases) f,pr glass stop-cocks. The simplest
lubricant is either pure vaseline or a mixture of pure vaseline
and resin .cerate.
A " soft" grease may be prepared by thoroughly mixing,
by melting and stirring, 1 part of beeswax and 3 parts of pure
vaseline. A" hard" grease is obtained by adding 1 part of
soft black rubber in small pieces to the latter mixture, heating
to 140 0 to 150°C, and stirring continuously until thoroughly
incorporated.
An excellent lubricant (Shepherd, 1931), which is particu-
larly suitable for gas analysis apparatus, is'l)btained from 6
parts of rubber (pale crepe, freshly milled, and free f~om dirt
and lint), 7 parts of white vaseline and 1 part of paraffin wax,
m. p. 40° to 45°C. The vaseline and paraffin wax are melted
in'a large porcelain casserole and the rubber is adqed in
several portions. The mixture is placed in a thermostated
oven or in an air bath (Section U, 12F; first paragraph)
and stirred continuously for 190 hours at 155 0 to 165°C. It
is then transferred to 2 oz. ointment jars and chilled in ice
immediately, care being taken to prevent condensation of
water on the grease.
'Excellent lubricants may be purchased. *
Calibration of a glass-stoppered burette. The burette is
filled with dichromate-sulphuric acid cleaning mixture and a
beaker placed beneath the burette. The burette is allowed
to stand for one hour, but preferably overnight. The solution
is run out and the burette rinsed five or six times with dis-
tilled water. The stop-cock and the inside of the ground joint
~re then wiped dry; the stop-cock is then smeared lightly
with a suitable lubricant, replaced into position, and turned
several times in order to spread the lubricant evenly in
the ground joint. The burette'is then clamped vertically and
filled with distilled water; the latter should. have been standing
for at least an hour prior to use so as to be certain that it has
acquired the room temperature. The temperature of the water
is noted. The air in the jet is expelled, by rUllI].ing out water
from the burette. Two methods are available Jor the actual
calibration. The first is based upon that employed at the
National Physical Laboratory. The second is due to Ost-
wald; it is very convenie~t to.use, but is not accepted by the
N.P.L.
Method 1. The burette is filled with water to a short
distance above the zero mark and water is swwly run out until
the meniscus is exactly on the zero mark. The drop of water
adhering to the i~t is removed by "bringing the jet into contact
with the side of the beaker. For reading the position of the
meniscus to 0·01 ml., the device shown in Fig. 2-21 coupled,
preferably, With a lens, may be employed. In the better
quality burette in which the calibrations are carried com-
pletely round for the 1 m!. divisions and half round for
fractions of a mI., errors due to parallax may be' avoided
by placing the eye so that the front and back of the
graduation at the meniscus appear to coincide. The burette
is allowed to deliver freely, i.e., with the stoIl-cock fully open,
into a clean, weighed Erlenmeyer (conical) fla.sk of 100-150
ml. capacity. When the meniscus of the water is approxi-
mately 1 cm. from the line to be tested, the rate of outflow
is reduced so that the motion -of the water surface is brought
undet complete control; the meniscus is adjusted exa9tly on
• A variety of lubricants is supplied by W. Edwards and Company. Vaushan
E9~d. London, 5.E.5, Ensland.
,
the mark.* No period of waiting for drainage should be
-allowed. The drop adhering to the jet after the setting has
been made is removed by bringing the side of the flask into
contact with the jet. T1:_J.e flask is then stoppered and
weighed to the nearest 0·01 g. This procedure is followed
for.each of the graduations to be tested, e.g., at 5 ml.. intervals,
5, 10, 15, 20, etc., or at 10 ml. intervals. The zero mark is
always taken as the starting point, since a very small drainage
error might take place between the readings and effect the
accuracy of the ,intermediate determinations. The whole
process should be repeated; the duplicate determinations
should agree within 2 centigrams. The volume actually
delivered is calculated from the weight .of the water and the
volume of 1 g. of water at the.temperature of the calibration
given in Table XXIX. The results should be tabulated as
follows.
CALIBRATION OF A BURETT¥: (TEMPERATURJt' = - 0C).
Burette readings Weight of True volume Correction

. (ml.) Water deliv- dt 20°C •(ml.)
e~d (mean)
0,00- 5·00
0,00-10,00
- g.
-g. . -m!.
-m!.
-m!.
-m!.
0·00-15·ob -g. -ml: -m!.
0·00-20'00, etc. -g. -m!. -mI.
.
The results may also be plotted graphically
with burette readings (1 m!. = 2 mm.) as
abscissae and corrections (0-01 m!. = 2 mm.)
as ordinates.
M ethfJd 2. The so-calleg. burette-calibrator is
employed. Thi~ consists of a small pipette
(usually of 2 ml. capacity)... with a mark on the
upper a:pd lower limbs, and a tap and side tube
fused to the lower end. I t is first thoroughly
cleaned and then attached by means of a short
piece of stout rubber tubing (" pressure tubing ")
to the jet .of the burette (Fig: 2-22); both
burette 'and pipette are clamped vertically:
The tap of the pipette is closed and distilled FIG. 2-22.
,
* If this has been missed by a small amount. the volume is read to the nearest
0·01 ml.
-
water is run in from the burette until its level is about
2 cm. above the upper mark. If all air bubbles have
not thus been removed from the pipette and its jet, more
water is run in and the adjust{llent repeated. The drop
of water adhering to the jet is removed by bringing
it into contact with the inside wall of a beaker. The jet is
then placed against the side of a tared weighing bottle or a
small stoppered flask, and water is allowed to flow slowly from
the burette-calibrator until the meniscus reaches the lower
mark; the bottle or flask is stoppered and reweighed. This
process is repeated at least six times, mId the mean weight
of the water delivered is obtained. The exact volume of the
pipette can then be calculated from the known temperature
of the water and Table XXIX.
The burette is now filled, adjusted exactly to the zero mark
by connexion with the pipette, and the latter, in turn, adjusted
to its lower mark. The pipette is then filled to the upper
mark from the burette, and the reading of the latter taken.
The pipette is emptied to the lower mark, and lihe process
repeated so that successive readings are taken at ca. 2, 4, 6, 8,
. . . . . 5() m!. marks on the -burette. From the known
volume of the burette-calibrator (say, 1·9962 mI.), the correct
volumes at the successive readings of the burette can be
computed and the corrections thus evaluated. The final
values may be collected in the form of a table or plotted
graphically as described under Method 1.
Standardised burettes. The National Physical Laboratory
recognises two types of tests for burettes, viz., " Class A "
and" Class B." For the former, only burettes with glass
taps are admitted: if a filling tube is sealed in between the
lowest graduation mark and the tap, a stop-cock must be
present in the filling tube.. Vario}ls details as to the construc-
tion of burettes are to be found in the pamphlet" Tests on
Volumetric Glassware': (N.P_L., April. 1934), but the toler-
ances and times of outflow (the 1atter are important in con-
,nexion with the drainage error) are tabulated below .
2
•
Total capacity, ml. 10 30 50 100 200
- - - -- - - - - -
Maximum error,
" Class A " .
± mi., 0-01 0·02 0·03 0'06 0·1 0·2
MaKimum error, ± mi.,

.. Class B" 0-02 0·04 0·06 0'1 0-2 0-4
The tolerances apply to the whole of the graduated portion
and to any portion of it; the differences in errors between
any two points must not exceed these values.
Length graduated, em. 20 30 40 50 60 70
Time
limits,
of outflow
seconds,
-- .
40-75 60-105 80-135 100-165 120-195 140-225
" Class A·"
Time of outflow
--
limits, seconds, 30-75 40-105 55-135 70-165 80-195 95-225
.. Class B"
, .
Only burettes which conform to .. Class A " tests are suit-
able for use in work of the highest accuracy and as ·standards.
The N.P.L. will, on payment of a certain f.ee, give a certificate
showing the exact volume delivered at various graduations
on " Class A " burettes. •
lit 23. Weight burettes.-For work demanding the highe~t
possible accuracy in transferring various quantities of
liquids, weight burettes are em-
J!loyed. As their name implies, they
are weighed before and after a
transfer of liquid. A very useful form
is shown 9-iagrammaticallyin Fig. 2-23,
a. There are two ground-glass caps, the
lower one is closed, whilst the upper
one is provided with a capillary open-
ing ; the loss by evaporation is
accordingly negligible. For hygroscopic
liquids, a small ground-glass cap is.
fitted to the top of the capillary tube.
The burette is roughly graduated in
5 ml. intervals.
The Lung-Rey pipette is shown in
Fig. 2-23, b. There is a small central
bulb (5 to 10 m!. c~pacity) closed by
two stop~cocks 1 and 2; the pipette 3
below the stop-cock has a capacity of
about 2 ml., and is fitted with a
(a) (b)
grountl.-on test tube 4. This pipette
is of particular value for the weighing FIG. 2-23.
out of corrosive and fuming liquids (see Section m, 12).
D, 24. Graduated cylinders.'"-These
are graduated vessels of thick glass (Fig.
2-24), and are available in capacities from
2 ml. to 2000 ml. Since the area of the
upper surface of the liquid is much
greater than in a measuring flask, the
accuracy is not very high. Graduated
cylinders cannot therefore be employed
for work demanding even a moderate
degree of 'accuracy. They are, how-
ever, useful where only rough measure-
. ments are required.· Two kinds are
generally employed: the wide-mouthed,
open type with spout (Fig. 2-24, A) and
(A) (B)
the stoppered type (Fig .. 2-24, B). They
may be calibrated, if desired, by measur-
FIG. 2-24 . .. ing info' 'them definite volumes of water
from calibrated butettes or pipettes.
,fi, 25. Storage and preservation of solutions in volumetric
analysis.-Solutions, which are comparatively stable and
unaffected by exposure to air, may be stored in whole or half
-winchester quart bottles; for work requiring the highelU:
accuracy Pyrex, or other r.esistance ~lass, bottles fitted with
ground-glass stoppers should be employed, the solvent action
of the solution being thus considerably reduced. Tohe bottle
should 'be clean and dry: a little of the stock solution is
introduced, the bottle well rinsed with thjs solution, drained,
and the remainder of the solution poured in and the b~ttle
immediately stoppered. If th.e bottle is not dry, but has
recently been thoroughly rinsed with tlistilled water, it may be
rinsed successively with three small portions of the solution
and'drained well after each rinsing; this procedure is, how-
ever, less satisfactory than that employing tne clean and dry
.vessel. Immediately after the solqtion has been transferred
to the stock bottle, it should be labelled with·(l) the name of
the solution, (2) its concentration, (3) the date of preparation,
and (4)' the initials of the person who prepared the solution,
together with any other relevant data. Unless th~ bottle is
c~pletely filled, internal evaporation and condensation. Will
cause dro~s of water to form on the upper part of the ,inside
• Also known as measuring cylinders.
of the vessel. For this reason, the bottle must be thoroughly
shaken before removing the stopper.
Solutions which are liable to alter their strength by access
of air (e.g., alkali hydroxides, ferrous ammonium sulphate,
titanous chloride or sulphate) may be kept in the apparatus
s how n diagrammati-
cally in Fig. 2-25,a. The
burette has a three-
-way tap which enables .
it either to be filled
from the stock bottle
or to be emptied. If
such a burette is not
available, one with a
sicie-tub~ and stop-cock
(Fig. 2-25, b) will serve
equally well. The tube
T is permanently can: (b)
nected with a source
of hydrogen (e.g., froin (a)
a Kipp's apparatus) if
the solution is oxidised
upon exposure to air
. (e.g., ferrous sul'phate, FIG. 2-25.
ferrous ammonium sulphate, titanous chloride or sulphate),
or to-a soda-lime guard tube, if it contains caustic alkali.
In the latter case, particularly if ordinary glass vessels are
used, the solution may become contaminated with silicates
owing to the attack of the alkali on the glass; it is Detter to
employ a ] ena glass storage vessel, or the inside of the vessel
may be coated with a thin layer of p3;raffin wax Qr with
" thermoprene."* .
An apparatus for the storage of large volumes of solutions
which alter their strength upon exposure to air, e.g., ferrous
sulphate solution, and which permits of the automatic filling
of the burette has been designed by Zintl and Reiniicker
(1926). A modification, due to G. F: Smith (1935), is shown
in Fig. 2-26. A is a large storage bottle of 10 to 15 litres
capacity. B is a 50 mI. burette provided with an automatic
filling device at C (the point of the drawn-out tube is adjusted
to be exactly at the zero mark of the burette), D is the
burette-bottle clamp, E is a two-holed rubber sto'):lper, F is a
• This is a. rubber paint made by the B. F. Goodrich Company, Akron,
Ohio, U.S.A.
256 Quantitative Inorganic ·A nalysis
ground glass tension joint, G is a rubber tube connected
to a source of hydrogen (for example, a Kipp's apparatus),
H is a Bunsen valve, and J is hydrogen. The burette is filled
l::JiI:===t====1m D
FIG. 2-26. FIG. 2-27.
by closing tap K and passing hydrogen through G; tap L is

then closed, tap K opened, and the excess. of l\quid allowed
to siphon back. .
(a)
FIG. 2-28.
Two other apparatus* for the storage of standard solutions
are shown in Figs. 2-27 and 2-28. These are adaptations of
mic~o-apparatus to the macro scale. Fig.,2-27 is self-explana-
tory. The solution is contained in the stoPage bottle A
and the 50 ml. burette is fitted into this by means of a gro.und-
glass joint B. To fill the' burette, tap C is opened and the
liquid pumped into the burette by means of the small bellows
E. F is a small guard tube; this is filled with soda-lime or
" sofnolite" when caustic alkali is contained in the storage
bottle. Bottles with '~. capacity up to 2 lit res are provIded
with standard ground-glass joints; large bottles, up to 15
litres capacity, can also be obtained. Fig. 2-28, a and b portrays
similar apparatus, but with an automatic filling device.
The solution is pumped into the burette and enters it through
a glass tube which terminates in a capillary exactly at the
zero mark; immediately the pressure is released, the solution
above the zero mark is automatically siphoned back into the
storage bottle.
• Obtainable from A. Gallenkamp and Co. Ltd., 17-29 Sun Street, Finsbury
Square, London, E.C.2, England.
TECHNIQUE OF GRAVIMETRIC ANALYSIS
the apparatus and operations <;ommon to both volumetric

and gravimetric analysis have been discussed in the preceding
Se<;tions, as has also the general theory unde'tlying gravi-
metric analysis. The experimental technique peculiar to
analysis by weight will be considered.in the following pages.
The operations of gravimetric analysis may be sum1llarised
under the headings: (a) precipitation, (b) filtration, (c) the
washing of the precipitate, and (d) the drying, ignition and
weighing of the precipitate.
fi, 26.. Precipitation.-The conditions for precipitation are

given in Section I, 59. Precipitations are usually carried out
in resistance glass beakers and the solution of the precipitant
is added slowly (for example, by means of a pipette, burette
or tap funnel) and with efficient stirring of the suitably diluted
solution. The addition must always be made without
splashing; this is best achieved by allowing the solution of
'the reagent to flow down the side of the beaker or precipitat-
ing vessel. Only a moderate excess of the reagent is generally
required; a very large excess may lead to increa~ing solubility
(compare Section I, 10) or contamination of the precipitate.
After the precipitate has settled, a few drops of the precipi-
tant should always be added ·to determine wheluer further
precipitation occurs. As a general rule, precipitates are not
filtered off immediately after they have been formed; most
precipitates, with the exception of those which are definitely
colloidal, such as ferric hydroxide, require more or less
digestion (Section I, 58) to complete the precipitation and
make all particles of filterable size. In some cases digestion
is carried out by setting the beaker aside and leaying the pre-
cipitate in contact with the mother liquor at room tempera-
ture for 12 to 24 hours; in others, where a higher temperature
is permissible, digestion is usually effected near the boiling
point of the solution. Hot plates, water baths, or even a low
flame if no bumping occurs, are employed for the latter pur-
pose; in all cases the beaker should be covered with a clock
glass with the convex side turned down or, if ayailable, with
a Fish,er " speedy-vap "beaker-cover. If the solubility of the
precipitate i,S appreciable, it. may be necessary to allow the
solution to alt,ain room temperature before filttation,
258
D, 27. Filtration.-This operation is the separation of thef'
precipitate from the mother liquor, the object being to get the
precipitate and the filtering medium quantitatively free from
the solution. The ;nedia employed for filtration are: (1)
filter paper; (2) filter mats of purified asbestos (Gooch
crucibles) or of platinum (Munroe crucibles); (3) porous
plates of glass (Jena sintered glass filtering crucibles), of
quartz (Jena sintered quartz filtering crucibles) or of porcelain
(porous porcelain filtering crucibles). The choice of the filter-
ing medium will be controlled by the nature of the precipitate
(filter p'aper is especially suitable for gelatinous precipitates
and for those which must be heated to a very high temperature
before weighing) and also by the question of cost. The latter
is of great importance in laboraiories where the number of
students is large and departmental grants small. The
limitations of the various filtering media are given in the
account which follow_s. ...
n, 28. Filter papers.-Quantitative filter papers must

have a very small ash content; this is achieved during manu-
facture by washing with hydrochloric and hydrofluoric acids.
The sizes generally used are circles of 7'0, 9'0, ll'O, and 12·5
cm. diameter, those of 9'0 and 11·0 cm. being most widely
employed. The ash of a 11 cm. circle should not exceed
0·0001 gram; if the ash exceeds this value, it should be
deducted from the weight of the ignited residue. Manufac-
turers give values for the average ash per paper: the, val~e
may also be determined, if desired, by igniting several filter
papers in a crucible. Quantitative filter paper is made of
various degrees of porosity. The filter paper used must be of
such texture' as to retain the smallest particles of precipitate
and yet permit of rapid filtration. Three textures are generally
nmqe, one for very fine precipitates, a second for the average
precipitate which contains medium sized particles, and a
third for gelatinous precipit_ates and coarse particles. The
~peed of filtration is slow for the first, fast for the thitd, and
•medium for the seeond. Hardened filter papers are made
by treating quantitative filter papers with nitric acid; these
have an extremely small ash, a much greater mechanical
strength when wet, and are more resistant to acids and alkalis.
They are strongly recommended for use in all quantitative
work.
The author h.as found "Whatman" filter papers very
satisfactory in practice; the characteristics of the chief
·varieties suitable for quantitative work are collected in the
following Table*; the ashes for the two most frequently
used sizes, 9 cm. and 11 em., are included.
TABLE XXX. "WHATMAN" QUANTITATIVE FILTER.-PAPERS.
Medium
Fast speed. speed. Slow speed.
Filler Paper. Retain Retain Retain
coarse l'IIedium- fine
particles. sized particles.
parti.cles.
Single acid (Hel) washed No. 31 No. 30 No. 32

Ash (mean) {l~ ~:: 0'00022 g.
0·00033 g.
0'00021 g.
0'00031 g.
0·00021 g.
0'00931 g.
Hardened single acid washed • No. 531
9·00013 g.
No. 530
0·00014 g.
--
Ash (mean) {9 cm.
_ 11 c.m 0'00019 g. 0·00021 g. -
Double acid (He], HF) washed'" No. 41 No. 40 No. 42
{ 9cm. 0·000102 g. 0·000090 g. 0'000060 g.
Ash (mean) 11 cm. 0'000l52 g. 0·000l34 g. 0'000089 g.'
Hardened dOll ble acid washed No. 541 No. 540 No. 542
{ 9 em. 0·000042 g. 0·000045 g. 0·000051 gJ
. 11 cm. 0'000062 g . 0·000067 g. 0'000076 g.
The filter papers Nos. 30-32 and 530-531 are tomparatively

cheap and may be used for elementary students. For precise
quantitative work, Nos. 40-42 or, preferably, the hardened
variety Nos. 540-542, which have a greater mechanical
strength and a greater resistance to acids and alkalis, should
be employed. Nos. 41 and 541 are used for gelatinous and
flocculen~ precipitates, and Nos. 40 and 540 for most other
. precipitates encountered in quantitative analysis.
The size of the filter paper selected for a particular operation
is determined by the bulk of the precipitate and not by the
volume of the liquid to' be filtered. The entire precipitate
should occupy about a third of the capacity of the filter at the
end of the filtration. As already pointed out, filter papers- of
9 cm. and 11 cm. diameter find the most frequent use.
A funnel with an allgle as nearly 60° as possible should be
employed; the stem should have a length of about 16 em.
in order to promote rapid filtration. The filter paper must be
carefully fitted into the funnel so that the upper portion beds
tightly against the glass. Some analysts recommend that
* Compiled from the pamphlet issued by the manufacturers. W. alld R.
Balston Ltd., Maidstone, England.
the filter paper should rest completely against the wall of the
funRel; this is really unnecessary since a filter paper which
adheres snugly to the funnel over the upper half will permit
•of more rapid filtration. To prepare the filter paper for use,
the dry paper is usually folded exactly in half and exactly
again in quarters. The folded paper is then opened so that a
60° angle cone is' formed with three thicknesses of paper on
the one side and a single thickness on the other; the paper is
then adjusted to fit the funnel. The paper is p,laced in the
funnel, moistened thoroughly with water, pressed down
tightly to the sides of the funnd either with the clean fore-
finger or with a flattened glass rod, and· then filled with water.
If the paper fits properly, the stem of the funnel will remain
filled with liquid during the filtration. Another method of
folding the filter which is preferable to that just described
3,0 to 4° for a 60° funnel); the corner of

~he fold should be torn off to a depth of
about one-third of the radius of the paper
(Fig. 2-29). When this filter is opened and
D
consists in folding the paper across a diameter and then once
again so that the two halves of the first crease do not quite
coil!cide (the two extreme edges should enclose an angle of
~
placed in the funnel, it should fit the walls
tightly at the upper half: if it does not fit
properly, the angle 6f the second fold must FIG. 2-29.
be adjusted until it does. Here also, the ultimate test of a

proper fit, is that the stem of the fminel remains fiijed with
liquid tht"oughout the p.ltration. It must be emphasised that
a funnel of proper size must be used: the
top of t!le filter paper when set in the funnel
should be about 1 cm. below the rim at the
latter.
To carry out a 'filtration, the funnel
containing the properly fitted paper is
'placed in a funnel stand (or is supported
vertically. in some other way) and a clean
beaker placed so that the stem of the
funnel just touches the side; this will pre-
vent splash,ing. The liquid to be filtered
is then poured down a glass rod into the
filter, directing the liquid against th,e side
of the filter and not in the apex (Fig. 2-30):
the lower end of the stirring rod should be
FIG. 2-30, very close to, but 'Should not quite touch,
the filter paper on the side having three thicknesses of
paper. The paper is llever filJed completely with the solu-
tion; the level of the liquid should not rise closer than to
within 5 to 10 mm. from the top of the paper. Any precipi-'
tate which adheres firmly to the side of the beaker or to the
stirring rod may be -removed with a rubber-tipped rod or
" policeman" (Se9tion n, 10L). •
Filtration may be appreciably accelerated by the use of the
lena .. analytical funnels" or the similar Pyrex 60°
funnels. These funnels have an exact angle of 60°, possess
a stem 100 to 150 mm. long with an internal diameter of
3 mm., and are specially grooved to permit of rapid filtration.
Filtration by suction with a filter paper (for apparatus, see
Section U, 31) is rarely necessary: with gelatinous and some
finely-divided precipitates, the suction will draw the particles
into the pores of the paper and the speed of filtration will
actually be reduced rather than increased. If suction is used
with filter paper, it is necessary to support the paper tn a
perforated cone made of platinum (" filter cone ") or in a
" Whatman " filter cone (hardened, No. 51)".
n, 29. Macerated filter paper.-Dittrich (1904) first recom-
mended the use of the filter paper pulp as an aid in the filtra-
tion and washing of gelatinous or slimy precipitates, which
tend to clog the pores of ordinary filter paper. The macerafed
paper may be prep.ared by vigorously shaking an ordinary
quantitative filter paper, torn into small pieces, with hot dis-
tilled water in a stoppered conical flask until it is disintegrated
to ,a pulp. This is unnecessary as macerated filter paper
tablets may be purchased ready for' use. Two .types are
avai1able: the first, known as Whatman " accelerators," is
in the form of oblong pieces (I" X til x -k"), eacli weighing
about 0·33 g. and giving an ash of about 0·000036 g.; the
second, known as Whatman " ashless tablets," is in the form
of discs (4·25 cm. diameter and about 4 mm."thickness), each
weighing about 2·28 g. and giving an ash of about 0·00025 g.
Either one or two accelerators or a quarter of a ta'Qlet is used
in the average precipitation; these disintegrate readily when
placed in water. The bulk of the macerated filter paper
should be approximately equal to that of the gelatinous
precipitate. The macerated filter paper is added afelr Ute
precipitate has formed and immediately before filtration.
When the bulk of the precipitate and paper is large, it is
usually advisa'Qle to support the filter paper on a Whatman
Technique ,()f Quantitative Analysis 263
filter cone (hardened, No. 51), and drain well on the pump upon
completion of the washing. It is best not to dry the filter
and precipitate completely as a h~rd mass m~y be formed,
which is difficult to insert in the crucible in the subsequent
ignition. It should be transferred while still slightly moist
to the crucible, and the drying completed in the crucible by
heating over a very small luminous flame. Alternatively, the
wet mass may, be ignited in a silica crucible, preferably
of the Main Smith form (see Section II, 35B).
Macerated filter paper is also useful when it is desired to
clarify a solution containing a finely divided precipitate tha1:
is difficult to coagulate.
II, 30. Filter mats. Gooch clllcibles.-F. A. Gooch (1878) ..
first employed a filter mat of purified asbestos; this was
supported inside a tall-form platinum crucible, the bojtom of
which was perforated "'with numerous small holes. Subse-
quently the idea was extended to porcelain crucibles, and in
recent years to silica crucibles. All crucibles of this type are
now termed Gooch crucibles. The porcelain type is the most
widely used for student and routine work. When the term
Gooch crucible is employed in this bo'qk, the porcelain variety
must be understood; crucibles of platimlm and of silica will
be calle,d platinum Gooch crucibles and silica Gooch crucibles
respectively. 'fhe asbestos must be carefully selected and
prepared: the long fibred, amphibole asbestos is the most
suitable. Gooch specified "white, silky, anhydrous asbes-
tos." The tedious process of preparing the asbestos by boiling
with hydrochloric acid, etc., is now usually carried out by
chemical manufacturers. "Asbestos for Gooch crucibles"
may be purchased at a reasonable price from any firm of
repute, and used with confidence. When SUGh asbestos is
shaken with water, it should separate into very small fibres;
no undisintegrated asbestos should be present. All {}ooch
crucibles are used under ,5uction, and require holders and
receivers that are strong enough to withstand a partial
vacuum. These are described in the following paragraph.
II, 31. Preparation of a Gooch crucible.-The Gooch crucible
(Fig. 2-31, A) is usually suppoited in a special holder, known as
a Gooch funnel, by means of a wide rubber tube (Fig. 2-31, B) ;
the bottom of the' crucible should be quite free from, the side
. of the funnel and from the rubber gasket, the latter in order
to be quite sure that the filtrate does not come into contact
with tpe rubber. The Gooch funnel passes through a one-
holed rubber bung into a large filter flask of about 750 niL
capacity. The tip of the funnel must pl'oject below the side
o (A) (6)
FIG. 2-31.
arm of the filter flask so that any risk that the liquid may be
sucked out of the filter flask may be avoided. The filter flask
should be coupled with another flask· of similar capacity, and
the latter connected with a water filter pump; if the water in
the.pump should" suck back," it will first enter the empty
flask and the filtrate will not be contaminated. I t is advisable
also to have some sort of pressure regulator to limit the
maximum pressure under which filtration is conduct~d. A
simple method is to insert a glass tap in the second filter
flas:k., as in Fig. 2-31, B; alternatively, a glass T piece may be
'introduced between the receiver and the pump, and one arm
closed either by a glass tap or by a piece of heavy rubber
tubing (" pressure" tubing) carry-
ing a screw clip. For some work
it is convenient to collect the 'filtrate
directly in a beaker: this may be
doqe by means of the apparatus
shown in Fig. 2-32. It consists of a
bell jar resting upon a ground-
glass plate and provided with tub-
ulures at the top and side, a three-
way tap is attached to the side tub-
ulure so that suction is easily applied
or stopped. Here also it is advisable
to provide a trap to protect the
FIG. 2-32. filtrate from the back flow from the
pump.
The Gooch crucible should be of suitable size. One about~
4 cm. in height, with a capacity of 25 mI., and perforations
about 0·5 to 0'8 mm. in diameter will be found serviceable
for most purposes. The crucible is first placed in the suction
filtering apparatus, and then half to two-thirds filled wit~ the
suspension of asbestos in water. The whole is allowed to
stand for 2 to 3 minutes in order to allow the larger particles
to settle to the bottom, and then suction is applied gently.
When the water has passed through, the pump is turned on
full, and the mat sucked down tight. The finaf uniform pad
of asbestos should have a thickness of 2 to 3 mm. (see Fig.
58, A); it is possible to tell approximately when the mat has
the correct thickness by holding the crucible up tq the light
and looking through it, wlten the 'outline of the holes should
be barely visible. If the pad is too thiI\. more asbestos must
be added and the procesS repeated. The asbest.os pad is now
thoroughly washed with distilled water under the maximum
suction of the pump until no fine fibres pass into the filtrate.
It should b~ mentioned that some analysts prefer to place a
perforated porcelain plate (ff Witt" plate, Fig. 2-31, C) upon
the asbestos mat to prevent its dislodgement: a little more
of the suspep.sion is poured in to furnish enough asbestos to
barely cover the plate and to hold it in place. This procedure
is unnecessary if it be remembered that no liquid may be
poured int~ the crucible unless suction is being-applie4. Liquids
should be poured gently on to the centre of the mat down a
stirring rod: a jet of water from a wash bottle should never
be directed into the prepared crucible. If these precautions
are taken there is little danger that the mat will become torn
and allow the precipitate to pass through.
The crucible is placed on'a small ignition dish or saucer or
upon a shallow-form vi~reosil capsu~e and dried to constant
weight at the same temperature as that which will be subse-
quently used in drying the precipitate. For temperatures up
. to about 250°C, a thermosfatically controlle.d electric oven
should be used. For higher temperatures, or if the crucible
requires ignition, the crucible may o
be heated in anelectrically-heaied
mume furnace. If the latter is
not available, it may be p'laced ..
inside a larger nickel crucible
fitted with an asbestos ring (as in
Fig. 2-33), and the larger crucible
heated directly with a flame. In
all cases the crucible is allowed to
cool in a desiccator (Section fl,
lON) before weighing. FIG. 2-33.
Attention is directed to vitreosil Gooch crucibles, which are
obtaipable in both translucent and transparent forms.
These are more expensive than the ordinary porcelain. cru-
cibles, but they possess many obvious advantages. They can
be heated to a much higher temperature without fear of
cracking, and are generally more resistant to chemical
reagents, particUlarly acids. If placed on a shallow vitreosil
capsule, they may be ignited directly over a flame.
n, 32. Munroe crucibles.-A platinum Gooch crucible, con-
taining a plati,num mat, was devised by C. E. Munroe in 1888.
The platinum mat is prepareg by igJriting a layer of ammonium
chloroplatinate, moistened with alcohol, in the crucible.
The chief features of such crucibles are: resistance to
chemical attack, rapidity of filtration, retention of the fi.nest
_precipitates, and possibility of heating to very high tempera-
tures. The platinum mat lasts almost indefinitely,. for the
residues can be removed by any solution treatment which
does not dissolve platinum. On account o~ ,the high price
of platinum, Munroe crucibles are rarely employed in routine,
analysis; they find some application for most exacting work,
as in atomic weight determinations.
n, 33. Crucibles fitted. with permanent porous plates.-
These possess am advantage over Gooch crucibles in so far that
no preparation of a filter mat is necessary. Tbe best known
ate the lena sintered glass crucibles. They are made of an
excellent resistance glass (lena Geraetglas 20, designated by
the symbol G "') .and have a porous disc
(I
\ - of sintered, ground gJass fused into the body

of the crucible (Fig. 2-34) .. The fused-in
£ritted filter disc can be obtained in various
degrees of porosity. Thus the average
size of the pores for plates numbered 1,
2, 3, and 4 are lQO-I20fL, 40-50fL. 20-30fL,
and 5-10fL respectively. No.4 is suitable
\\ II for very fine precipitates (such as barium
sulphate) ;md No. 3 .for precipitates of
FIG. 2-3-1.
medium particle size. Two types of crucible
find the widest application in quantitative analysis, the
low wide form (designated by the number 10: thus 10G3
would refer to such a crucible with porosity 3) and the
higher or tall form (designated by the number 1, e.g., IG4).
A crucible with a capacity of 30 ml. is satisfactory for most
work.
The advantages of Jena sintered glass crucibles are:" (i)
they are made eritirely of glass, which is resistant to most
chemical reagents with the exception of hydrofluoric acid and
hot concentrated alkalis; (ii) they can be dried to constant
weight at 100°-150°C; and (iii) they are readily cleaned.
These crucibles can be heated to a temperature not exceeding
500°-550°C; if heating above 150°C is required, they should
be placed in a cold electric furnace, heated to·the necessary
temperature, and all.owed to cool to about 200°C in the fur- •
nace before being r~moved to a desiccator.
For temperatures above 500°C, lena quartz filter crucibles
may be employed. These are similar to the J ena sintered
glass crucibles, but are made of pure fused silica. They can
be used up to 1300 0 e, and there is no danger of cracking
through sudden changes of temperature. Two types are
made, designated by the symbols IbB3 and 4 and by 10B3
and 4. They are attacked by phosphates and strong alkalis,
but are otherwise very resistant to chemical reagents.
Another vari~ty of crucible, which is likely to 'find wide
application when its properties are better known, is the
porous porcelain filter crucible. It consists of a porcelain
~rucible, glazed-inside and outside, with a porous porcelain
bottom; the latter is available in a number of porosities.
The crucible may be heated to bright redness in-a furnace or
it may be heated inside a larger metal crucible (Fig. 60)
directly with a burner. Heating and cooling should be
gradual in order to minimise the danger of cracking.
Crucibles fitted with permanent porous plates are cleaned
by shaking out as much of the solid as possible, and then
dissolving out the remainder of the solid with a suitable solvent.
An alternative method is to suck water or
some other 'liquid through from the reverse
side. J:he process will be evident upon
reference to Fig. 2-35.
n, 34. Washing of precipitates.-Most
precipitates are produced in the presence of
one or more soluble compounds. Since the
latter are frequently not volatile at the tem-
perature at which the precipitate is ultimately
dried, it is necessary to wash the precipitate
to remove such material as completely as FIG. 2-35 ....
possible. The minimum volume of the waship.g liquid
required to remove the objectionable matter should be used
since no precipitate' is absolutely insoluble. ~ualitative
"268 Quantitative Inorganic Analysis
tests for the removal of the' impurities should be made on
small volumes of the filtered washing solution. Further-
more, it is better to wash with a number of small portions of the
washing liquid, which are well ,drained between each washing,
than with one or two large portions, or by adding fresh por-
tions of the washing liquid whilst solution still remains on the
filter (see Section I, 60). The precipitate should never more
than one-third to one-half fill the filter paper, and the liquid
should be kept from 5 to 10 mm, below the edge,
The ideal washing liquid should comply as far as possible
with the following conditions :
(I) it should have no solvent action upon the precipitate,
and dissolve foreign substapces easily;
(2) it should have no dispersive action on the precipitate;
(3) it should form no volatile or insoluble product with the
precipitate;
(4) it should be easily volatile aHhe temperature of drying
of the' precipitate;
(5) it should contain no substance which is likely to inter-
fere with subsequent determinations in the filtrate.
In general, water should not be used unless+it is certain that
it will not dissolve appreciable amounts of the precipitate or
peptise it. -If the precipitate is appreciably soluble in water,
a common ion is usually added, since any electrolyte is less
soluble in a dilute solution containing one of its ions than it is
in pure water (Section I, 9); as an example the washing of
calcium oxalate with dilute ammonium oxala_te so,lutioh may
be cited. If the precipitate tends to become colloidal and
pass through the filter paper (this is frequently observed with
gelatinous or flocculent precipitates), a wash solution contain-
ing an electrolyte must be employed (comp~re Section I, 56).
The nature of the electrolyte is immaterial provided iL has no
action upon the precipitate during washing and is volatilised
during the final .heating. AIl\monium salts are usually
selected for this purpose: thus ammonium nitrate solution is
employed for washing ferric hydroxide. In some cases it is
possible to select a solution whicJ:t will both reduce the solu-
bility of the precipitate and prevent peptisation, for. exa_mple,
the use of dilute nitric acid with silver chloride. Some pre-
cipitates tend to oxidise during washing; in 'such instances
the precipitate cannot be allowed to run ~y, and a special
washing solution which reconvert~ the oxidised compounds
into the original condition must be employed, e.g., acidulated
hydrogen sulphide water for copper sulphide. Gelatinous
precipitates, like aluminium hydroxide, require more washing
than cry~talliiie ones, such as calcium oxalate. With gel a- ~
tinous precipitates there is also a danger of channel formation
if the wash liquid is allowed to drain completely; these .pre-
cipitates should be washed as far as possible by decantation.
The reader may have observed that full dGtails of the
technique of filtration have not so far been given. This is
because the filtration of a precipitate is. so closely associated
with the washing process that a consideration of the former is
logically deferred until the latter has been discussed, at least
in outline.
II, 35. Technique of filtration.-When. the proper filtering

medium (filter paper, Gooch crucible, etc.) has been prepared,
as much- as possible of the supernatant liquid is poured off
without disturbing the precipitate by directing the stream of
liquid against a glass rod held against the lip of the beaker
(compare Fig. 2-30). The precautions already mentioned
against filling a filter paper too full must be taken. In most
cases, particularly if the precipitate settles. rapidly or is
gelatinous, washing by decantation may be employed. Twen.ty
to fifty m}. of a suitable wash liquid is added to the residue in
the beaker, the solid stirred up and allowed to settle. If the
solubility of the precipitate allows, the solution should be
heated, since iDter alia the rate of filtration will thus be
increased. When the supernatant liquid is clear, as much as
possible of the clear liquid is decanted through the filtering
medium. This process is'repeated three to five times (or as
many times as is necessary) before the precipitate is trans-
ferred to the filter. The main bulk of the precipitate is first
transferred by mixing with the wash solution and pouring
off the suspension, the process being repeated until most of the
solid has been removed from the beaker. The precipitate
adhering to the sides and the bottom of the beaker is removed
as follows. The beaker is grasped in the left hand and the
stirring rod is held firmly against the top of the beaker with
the index finger and 'Should project 2 to 3 cin. beyond the
lip; the wash bottle is controlred by the right hand. The
beaker is inclined and a stream of water (or wash liquid) is
directed against the precipitate to dislodge it and wash it
against the rod into the filter. The process is shown diagram-
matically for a filter paper in Fig. 2-36. For some precipitates,
...where washing by decantation is inadmissible because-of
solubility influences, the whole-of the precipitate must be
transferred to the filter in this way. After the above treat-
_. ment there will generally be small amounts of the precipitate
adhering to the walls of the beaker.
These are removed by rubbing with a
" policeman " (Section II, lOL); when
all the particles have been dislodged,
the "policeman" is rinsed with the
wash liquid, and the remaining precipi-
tate transferred to the filter.
Where the precipitate is washed 'on
the filter, .in the last stages the washing
solution is directed along the rim and
then gradually towards the apex of the
coI}~. In all cases, tests for the com-
ple~el'less of washing must be made by
collecting a small sample of the washing
FIG. 2-36. . solution after it is estimated that most
pf the impurities have been removed, and
applying an appropriate qualitative test. Where filtration
is carried out under suction, a small test tube may be
at.tached to the bottom of the Gooch funnel by means of a
WIre. .. ;
II, 36. Drying and ignition of precipitates.-After a predpi-

tate has been filtered and washed, it must be brought ~o a
constant cqmposition before it can be weighed. The further
treatment will depend both upon the nature of the precipitate
and upon that ot the filtering medium. This treatment
consists in drying or igniting the precipitate. Which of the
latter two terms is employed depends upon the temperature
at which the precipitate is heated. There' is, 'how~ver, no
definite temperature below or aMve' which the precipitate
is said to be dried or ignited respectively. The meaning will
be adequately conveyed for our purpose if we designate
drying wl1en the temperature is below 250°C (the maximum
temperature ~hich is readily reached in the usual thermo-
statically-controlled, electric drying oven), and ignition above
250°C up to, say, 1200°C. Precipitates that are to be dried
should be collected on porcelain,or silica Gooch, Jena si9-tered
glass or porous porcelain filter crucibles. PrecipitateS that
are to be ignited are collected on filter paper, porcelain, but
preferably, silica ~ooch crucibles, porous porcelain filter
crucibles, Jena sintered glass crucibles (not above 550 q, 0
Jena quartz filter crucibles, or, less commonly, in platinum
Gooch crucibles. Ignition in crucibles containing either
asbestos or a sintered glass disc as the filtering medium is
simply effected 'by placing in a larger nickel or platinum'
crucible, as in Fig. 2-33, and heating with the appropriate
'lmrner (Section U, lOA); alternatively, these crucibles (and,
indeed, any type of crucible) may be placed in an electrically-
heated muffle furnace~ which is equipped with a pyrometer
and a means for controlling the temperature.
Many of the precipitates which are subsequently to be
ignited are filtered through filter paper, and it is therefore
necessary to describe the method to be adopted in such cases.
The exact technique will depend upon whether the precipitate
may be safely ignited-in contact with the filter paJ2er or noL
It must be remembered that some precipitates, such as barium
sulphate, may be reduced 6r changed in contact with ·.filter
paper or: its decomposition products.
A. Incineration of the filter paper in the presence of the-precipitate.
On account of cheapness, porcelain crucibles are usually
employed fo):" this operation, particularly in instructional
courses. Silica crucibles. although somewhat more expen-
sive, are to be preferred because of their greater resist-
ance to thermal shock. The crucible is first ignited to con-
stant weight (i.e., to within 0'0002 g.) at the same temperature
as that to which the precipitate is
ultimately heated. The well-drained
filter paper and precipitate are carefully
detached from the funnel; the filter paper
is folded so as to completely enclose the
precipitate, care Being taken not to tear
the paper. * The packet is then placed t
point down in the weighed crucible, which
is supported <;>n a pipe-clay or, better, a
silica triangle resting on a ring stand
as in Fig. 2-37.t The crucible is slightly
inclined, as -shown in the diagram, and
covered partially with the lid, which
• should rest partly on the triangle. A very FIG. 2-37.
~
• For bulky. preCipitates, such as those containing macerated filter paper, it

is advisable to remove most of the..mois'ture by a preliminary drying in the
steam oven, or at lOO-105°C.
t It is preferable to attach the triangle to the ring of a retort stand by bend-
ing the wire ends of the triangle round it rather than to support it on a tripod.
In the former case, the crucible is more readily accessible and also the height
of the crucible above the burner can be easily adjusted.
small flame is then placed. under the crucible lid; drying thus
proceeds quickly and without undue risk. When the mois-
ture has been expelled, the flame is increased slightly so as to
'slowly carbonise the paper. The paper should not be allowed
to inflame as this may cause a mechanical expulsion of fine
particles of the precipitate owing to the rapid escape of the
products of ~ombustion; if, by chance, it does catch fire, the
flame should ·be ex~ingu~shed by momentarily placing the
cover on toe mouth of the crucible with the aid of a pair of
crucible tongs. When the paper has completely carbonised
and vapours are no longer evolved, the flame is moved to the
back (bottom) of the crucible and the carbon slowly burned
off whilst the flame is slowly increased. * After all the carbon
has been burned away, the crucible is covered completely (if
desired, th·e crucible may iJe placed in a vertical position for
this purpose) and heated to the' required temperature by
means 1)f a Bunsen, Mcker or Fisher flame. Usually it takes
about 20- minutes to char the paper, and 30 to 60 minutes to
compl~te the ignition.
When the ignition is ended, the flame is removed and, after
1 to 2 minutes, the crucible and lid are placed in a desiccator
cofltaining a suitable desiccant (Section fi, 10 N), and allowed
to cool for 25 to 30 minutes. The crucible and lid are·thl:fIl
weighed. The crucible and contents are then ignited at the
same temperature for 10 to 20 minutes, allowed to cool in a
desiccator as before, and weighed again. The ignition, is
repeated until constant weight is attained. Crucibles should
always be handled with clean crucible tongs.
·B. Use of crucible with interior-fitting serrated lid. J. D.
Main Smith (1925) has designed an improved form of crucible
lid. This is an interior-fitting lid of which the vertical flange
is serrated around the edge; it fits
into a squat-shaped crucible, the
upper surface of the lid being about 1
cm. from the top of the crucible (Fig.
2-38). The advantages are: (i) gases
or vapours, which are disengaged in the
operation, can escape freely; (ii) loss
FIG. 2-!l!!. by creeping, spattering and decrepita-
tion are prevented; (iii) the contents of the crucible can be
$:!asily inspected, since the lid easily swivels up by pressing on
• If the carbon on the lid is oxidised only slowly, the cover may be heated
separately in a flame. It is of course held in clean crucible tongs.
T~chnique of Qu.antitative Analysis 273
one side lightly with the clean crucible tongs (this is, of course,
unnecessary if the lid is of transparent silica); (iv) liquids
can easily be added simply by pouring on to -the lid, the
serrations providing ingress into tqe crucible. Vitreosil
crucibles of 15, 25, 40, and 50 m!. capacity with serrated lids
are commercially available, and are comparatively inexpen-
sive. Lids of transparent silica can also be obtained.
The ignition of a precipitate in a wet filter paper is readily
effected by jJIacing'the precipitate and filter into a previously
ignited and weighed crucible and lid, .and conducting the
initial heating exactly as described under A. Care must
be taken that a sufficiently large crucible is used (usually
of 40 m!. capacity); an over:large crucible is no disadvan-
tage. After the paper has been carbonised and vapours
are no longer evolved, the lid is removed and laid aside
upside-down on a clock glass, and the combustion of the
carbon completed at a red heat in the usual manner. If the
lid has been blackened by the preliminary heating, it should be
held by the ring upside-down by means of clean crucible
tongs, and carefully ignited over the open crucible by means
of a small flame until all the carbon has been burned off; the
. clean lid is then replaced in the crucible, care being taken that
any.fragments of residue present fall into the crucible. The
crucible and lid are then ign.ited at the appr6priate tempera-
ture to constant weight.
C. Incineration of the filter paper apart from the precipitate.
This .method is employed in all those cases· where the ignited
substance is reduced by the burning paper, for example,
barium sulphate, lead sulphate, bismuth oxide, copper oxide,
etc. The funnel containing the precipitate is covered by a
piece of qualitative filter paper upon which is written the
formula of the precipitate and the name of the owner; the
paper is made secure by crumpling its edges over the rim of
the funnel so that they will engage the outer conical portion
of the funnel. The funnel is placed in the steam oven, or in
a drying oven maintained at'100° to 105 0 C, for 1 to 2 hours or
until completely dry. A sheet- of glazed paper about 25 cm.
square {white or olack,. to contrast with the' colour of the
precipitate) is placed on the bench away from all draughts.
The dried filter is removed from the funnel, and as much as
possible of the precipitate is removt;d from the 'paper and
allowed to drop on a clock glass resting upon the glazed paper.-
This is readily done by very gently rubbing the sides of the
filter paper together, when the bulk of the precipitate becomes
detached and drops ·upon the clock glass. Any, small particles
of the precipitate which may have fallen upon the glazed paper
are brushed· into the crucible with a small camel-hair brush.
The clock glass contaIning the precipitate is then covered
with a larger clock glass or with a beaker. . The filter paper is
now carefully folded and placed inside a weighed porcelain or
silica crucible. The crucible is placed on a triangle and the
filter paper incinerated as detailed in the first paragraph of this
Section. The crucible is allowed to cool, and the filter ash
subjected to a suitable chemical treatment in order to convert
a"ny reduced or changed material into, the form finally desired.
The cold crucible is then placed upon the glazed paper and the
main part of the precipitate carefully transferred from the
clock glass to the crucible. A small camel-hair brush will
aSSIst in the transfer. Finally, the precipitate is brought to
constant weight by heating to the necessary temperature.
Here also the Main Smith crucible may be used. The
drjed filter paper, from which most of the pre9ipitate has been
detached as described above, is first ignited exactly as already
detailed. The crucible and lid are allowed to cool, a small
quantity of the liquid chemical required to convert any re-
duced or changed tnaterial to the form finally desired is poured
on to the lid, and the whole sl9wly heated to the correct
temperature. 'the crucible is then allowed ,to cool, placed
upon a sheet of glazed paper, the main part of the precipitate
added, and the whole ignited to constant weight at the appro-
priate temperature.
n, 37. The care.and use of plathtum vessels.-Platinum has
a very high melting point (1770°C). The pure metal is, how-
ever, too soft for general use, and is therefore always hardened
with small quantities of rhodium, iridium, or gold. These
alloys are slightly volatile at temperp.tures above llOO°C,
but retain most of the advantageous properties .of pure
platinum, such as resistance to most chemical reagents
including molten alkali carbonates and hydrofluoric acid (the
exceptions are dealt with below), excellent conductivity of
heat, and extremely small adsorption of 'water vapour. If
the financial resources of the laboratory are limited, at least
two crucibles of 25 :rnL capacity, provided with covers, should
be purchased; these will meet most ordinary requirements.
It is useful, if finances permit, to have crucibles of 10 mL
capacity and platinum dishes of 100, 300 and 600 ml. capacity.
Platinum crucibles should be supported when heated upon
a platinum triangle. If the latter is not available, a silica or
pipe-clay triangle may be used;_ nichrome and other metal
triangles should be avoid~d. Hot platinum crucibles must
always be handled with platinum-tipped crucible tongs, or
with ordinary tongs which are cpvered with platinum foil,
securely attached by platinum wire. Unprotected brass or
iron tongs produce stains on the crucible. Vessels of pla~
num must not b~ exposed to a luminous flame, nor may they
o
I
'ncorrect position Correct position Correct position
Bunsen flame Meker or fIsher flam,
FIG. 2-39.
be allowed to come into contact with the inner cone of a gas

flame (see Fig. 2-39); t_hismay result in thedisintegrationofthe
surface of the metal, causing it to become brittle, owing,
probably, to the formation of a carbide of platinum.
Platinum apparatus may be used without sensible loss for* :
1. Fusions with (a) sodium carbonate or fusion mixture, (b)
borax, (c) alkali bifluorides, and (d) alkali bisulphates (slight
attack in the last case above 700°C, which is diminished by the
addition of ammonium s-qlphate).
2. Evaporations with (a) hydrofluoric acid, (b) hydrochloric
acid in the absence of oxid~sing agents which yield nascent
chlorine, :lnd (c) concentrated sulphuric acid (a slight attack
may occur). ,
3. Ignition of (a) barium sulphate and sulphates of metals
which are not readily reducible, (b) alkaline earth carbonates,
oxalates, ·etc., and (c) oxides which are not readily reducible,
e.g., CaO, SrO, BaO, A120 3 , Cr20 3 , Mn a0 4, Ti02, Zr0 2 , Th02 ,
MOO-a, and W0 3 •
• Extracted from the booklet entitled Platinum Labaratory Apparatus
issued by Johnson Matthey and Co. Ltd., Hatton Garden, London, E.C.I,
England. •Further qetails will be.tSmnd in this booklet.
Platinum is attacked under the following cOI).ditions, and
such operations must -not be conduded in platinum vessels:
1. Heating with the following liquids: (a) aqua regia,
(b) hydrochloric acid and oxidising agents, (c) liquid mixtures
which evolve bromine or iodine, (d) concentrated sulphuric
acid (slight action, which -is, however, negligible for most
purposes).. and (8) concentrated phosphoric acid (slight, but
appreciable ~ction after prolonged heating).
2. Heating with the following solids, their fusions or
vapours: (a) oxides, peroxides, hydroxides, nitrates, nitrites,
sulphides, cyanides, ferricyanides and ferrocyanides of the
alkali and alkaline earth metals, (b) molten lead, silver, copper,
bismuth, tin or gold, or mixtures which form these metals
upon reduction, (c) phosphorus, arsenic or antimony, or
mixtures which form these elements on reduction, particu-
larly pllosphates in the presence of reducing agents, (d)
sulphur (slight action), selenium, and tellurium, (e) volatile
halides, especially those which readily 'decompose, and (f)
substances of unknown composition.
Cleaning and preservation of platinum ware. All platinum

apparatus (crucibles, dishes, etc.) should be kept clean,
polished and in proper shape. If, say, a platinum crucible
becomes stained, it should be fused with a little sodium car-
bonate, the-molten solid poured out on to a dry stone or 'iron
slab, the residual solid dissolved out with water, and the
vessel then digested with concentrated hydrochloric acid:
this treatment may be repeated, if necessary. If fusion with
sodium carbonate is without effect, potassium hydrogen
sulphate may be substituted; a slight attack of the platinum
will occur. In some cases, the use of hydrofluoric acid or
potassium hydrogen fluoride may be necessa~y.
If the surface exhibits signs of becoming dull, the vessel
must be polished with fine silver,sand, say of 75 to" 100 mesh.
Care must be taken that the sand consists of rounded grains
(examine under microscope) and that it is free from angular
pieces and grit, which would scratch the metal. The sand
should be moistened and applied ~ith the finger or a soft
cloth.
All platinum vessels must be handled with care to prevent
deformation and denting. Platinum crucibles must on no
account be squeezed with the object of loosening the solidified
cake after a fusion. Boxwood for,m~rs can pe purchased for
Techni'fJue of Quantitative' Analysis 277
crucibles and dishes; these are invaluable for reshaping
dented or deformed platinum w~re,
II, 38.· Perforated screens for crucibles.-It is often impor-
tant to exclude flame gases from the interior of a crucible during
an ignition, e.g., in the ignition of ferric oxide'. For this
purpose we may employ-a vitreosil plate, about 10 cm. square,
in which a round opening is cut large enough to admit the
crucible to two-thirds of its depth. The plate is held at an
angle of about 30° from the horizontal by means of a clamp;
,alternatively, but less satisfactorily, it may be suspended on a
tripo"'d. Asbestos board may also be employ~d, but this has
the disadvantage that fibres may adhere to the crucible:
this difficulty is less likely to occur with" uralite." To ob-
tain the maximum temperature, the following device may be
used. An asbestos, " uralite " or vitreosil sheet is provided
with a comparatively large hofe, say, 5 em. in diameter;
upon this is placed a thin sheet of platinum with an, opening
just large enough to support the. crucible.
CHAPTER III
'VOLUMETRIC
~
AN ALYSIS
ACIDIMETRY AM> ALKALIMETRY
Ill, 1. PrepartitiOD of a standard acid. Discussion.-Two
acids, namely, hydrochloric acid and- sulphuric acid, are
widely employ~d in the preparation of standard solutions of
acids. Both .o~ these are commercially avp.ilabie as concen-
trated solutions; concentrated hydrochloric acid is about
10·5 to 12N, and concentrated SUlphuric acid is about 36N.
By suitable dilution, solutions of any desired approximate
strength' may be read.ily- prepared. HydrochlQric acid is
generally more convenient, since most chlorides are soluble in
water. Sulphuric acid forms insoluble salts with lime and
barytq; for titration of not liquids, or ..for determinations
which require boiling for some time with excess of acid, stan-
dard sulphuric acid is, however, preferable. Nitric acid is
rarely employed, because it almost invariably contains a
little nitrous acid, which has a destructive actiQn upon many
indicators.
For the present, we shall confine our attention to the
preparation of standard solutions of hydrochloric acid. Two
methods are available. The first utilises the experimental
fact that aqueous solutions of hydrochloric acid lose either
hydrogen chloride or water upon boiling, according as to
whether they are stronger or weaker, until they attain a
practically constant composition (constant boiling point
mixture), which depends upon the prevailing pressure. The
composition of this constant boiling mixture and its depend-
ence upon pressure has heen determined with great accuracy
by Hulett and Bonnet (1909) and by Foulk and Hollings-
worth (1923). The relevant data are collected in Table XXX.
The constant boiling point acid is neither hygroscopic nor
appreciably volatile, -and its concentration remains unc'\langed
if kept in a well-stoppered vessel out of direct sunlight.
This acid may be employed directly in the preparation of a
solution of hydrochloric acid of known concentration.
In the second method a solution of the approximate strength
required is prepared, and this is standardised against some
standard alkaline substance, such as borax or anhydrous
278
V olttmetric Analysis 279
TABLE XXX. COMPOSITION OF CONSTANT BOILING POINT

HYDROCHLORIC ACID
-
.
Gr.ams# of Acid,
Pressure (mm. of I Per cent HCl in weighed in air,
Hg) Acid (vac. wt.) contaIning 36·47 g.
of HC1 .
.
780 20·173 180·621
770 20·197 180·407
760 20·221 180·193
750 20·245 179·979
740 . 20·269
20·293
179·766
730 179·555 -
. •
.odium carbonate; other suitable substances are given in
,ection ITI, '1. Both methods will be."described. If a solution
of an exact normality is required, a solution of an approximate
strength somewhat greater than tha~ desired is first prepared;
this is suitably diluted with water after standardisation (for
a typical calculation, see Section 1,24, Example 16).
The student should read the following theoretical sections
bejore embarking upon, the experimental work: .
Sections I, 27 to I, 29-theory of acidimetry and alkali-
metry;
Section I, 23-equivalent weights, normal solutions.
m, 2. Preparation of constant boi1.iu.g point hydrochloric
acid._:Method A (rA. Sha:w, 1926). Mix 500 ml. of pure con-
centrated hydrochloric acid (sp. gr. 1'19), preferably of
analytical reagent quality, with 500 ml. .of distilled water* ;
the resultant acid sbould have a sp~cific\ gravity of 1'10
(test with a hydrometer). Fit up an apparatus consisting of
a 1·5 litre Pyrex distilling 'flask surmounted with a splash
bulb but no trap (which might set up a back pressure), and
connected to a Liebig's condenser having a 75 cm. water
jacket and an inner tube of 6 to 10 mm. bore. Insert the end
of the condenser inside the neck of a, Pyrex Erlenmeyer
(conical) flask. Make.all connexions with new rubber stop-
pers, and arrange that the glass tubing projects at least 2
cm. beyond tlfe rubber. Place the diluted acid in the flask.
To prevent superheating, allow the flask to test on a sheet of
• • If the concentrated acid has a specific gravity of 1'16, as is often the case,
300 ml. of water are required for 500 ml. of acid.
280 Quantitative· Inorganic Analysis
asbestos in which a hole is cut for the flame; insulate the
upper part of the flask against heat radiation by means of
asbestos paper or cloth. CQnduct the distillation continu-
ously at a rate not exceeding 8 to 10 ml. per-minute. Do not
immerse the ~ondenser tip in the distll1ate, but allow the
condensate to flow freely into the flask. Set aside the first 75.
per cent of the volume of the original distillate for ordinary
laboratory use, and retain the next 1l) to 15 per cent as con-
stant boiling point acid. Discard the residue in the flask
(about 100 mI.). Read the barometric pressure to the nearest
mm. at the beginning of the distillation, agaill after 75 per
cent has passed over, and finally at the end of the distillation .. '
The mean of the last two readings is calculated and the
concentration of the acid interpolated from Table XXX.
Meth·od B. Mix 400 m,l. of pure concentrated hydrochloric
acid with 250 to' 400 mI. of distilled water so that the specific
gravity of the resultant acid is 1,10 (test with a hydrometer).
Insert a thermometer in ''the neck of a 1 litre Pyrex distilling
flask so that the bulb is just opposite ,tM side tube, and
attach a Liebig's condenser to the side tube; use new rubber
bungs. Place 500 m!. of the diluted acid in the flask and distil
the liquid at a rate not exceeding 8 to 10 m!. per minu,te (3 to 4
ml.. per minute is a convenient rate), and collect the distillate
in a small Pyrex flask. From time t6 time pour -Dff the dis-
tillate into a .500 ml. measuring cylinder. When 375 ml,. has
been collected and transf~rred to the measuring cylinder,
collect a further 50 ml. in the small Pyrex flask; watch the
thermometer to see' that the temperature remains constant.
Remove the receiver and cork it. This contains the pure
consjant boiling point acid. Note the barometric pressure
to the nearest mm. during the distillation and at the end, and
take the mean value. _ Interpolate the c.oncentration of the
acid from Table XXX.
m, 3. Direct preparation of O'IN hydrochloric acid from
the· constant boiling ·point acid.-Weigh a small, dry conical
flask, fitted with a cork. Add the calcUlated quantity of
constant bojling point acid required for the preparation of 1
litre of O'lN acid (see Table XXX) with the aid of a pipette;
make the final adjustment with a small piece of glass capil-
lai'Y, to one end of 'which a piece of rubber tubing is attached.
By this method it should be possible to weigh the acid to the
nearest centigram. Reweigh the flask to 0·061 g. after
replacing the stopper. Add-an equal volume of water to
V.olumetric Analysis 281
prevent loss of acid, and transfer the contents to -a litre volu-
metric flask. Wash out the weighing flask several times with
distilled water and add. the washings to the original solution.
Make up to the mark with distilled water. Insert the stopper
<and mix the solution thoroughly by sha~ing and inverting the
flask repeatedly.
No'ttl.-Unless a solutjon of exact normality is required, ·it is not
necessary to weigh out the exact quantify of constant boiling acid.
The exact normality may be calculated from the ::.veight of- acid
used. Thus, if 18·305 g. of acid, prepared at 760 mm., was diluted
to 1 litre, its normality would be 18·305/180·193 = 0·lOU8N.
Solutions of other concentrations are similarly prepared.
m. 4. Preparation of 0 ·1N hydrochloric acid and standardi-
sation.-Measure out by means of a graduated cylinder or a
burette 9 mI. of pure concentrated hydrochloric acid; pour
the· acid into a litre volumetric flask or into a litre measuring
cylinder (Fig. 2-24B) containing about 500 ml. of distilled water.
Make up to the litre mark with distilled water and thoroughly
mix by shaking. This will giv.e a solution approximately
O'IN (1).
Note. 1. If IN hydrochloric acid is required, use 90 m!. of the
conc~ntrated acid. If O·OIN acid is' required, dilute two 50 ml.
portions of the approximately O'IN acid, removed with a 50 m!.
pipette, in a volumetric flask to 1 litre.
Approximately O·IN sulphuric acid is similarly prepared from 3
ml. of pure concentrated sulphuric a~id.
Two excellent methods are available for standardisatioJ;l.
ihe first has beep widely employed, but the second, of more
recent date, is more convenient, less time-consuming, and
equally accurate.
A. Standardisation with anhydrous sodium carbonate.
Pure sodtum carbonate. Analytical reagent quality* sodium
carbonate of 99·9 per cent purity is obtainable commercially:
This ,contains a little moisture. and must be dehydrated by
heating at 260° to 270°C for half an hour and allowed to cool
in a desiccator before use. Alternatively, pure sodium
carbonate may be prepared by heating A.R. sodium bicar-
bonate to 260° to 270°C for 60 to 90 minutes; the temperature
must not be allowed to exceed 270°C, for above this tempera-
ture the sodium carbonate may lose carbon dioxide. A small
quantity (5 to 10 g.) of A.R. sodium bicarbonate is placed in
a porcelain, nickel or, best, a platinum 'crucible, and pressed
• Analytical reagent will be abbreviated to A.R..
282 Quantitative Inorganic .A nalysis
against the walls pf the crucible so as to form a uniformly
thick layer. The crucible is heated by any of the following
methods:
(i) Embed the cruc,ible in a sand bath, immerse the bulb
of a 360°C thermometer in the sand adjacent to the crucible,
and heat the sand bath with, a .fre~ flame; maintain the
temperature at ,260° to 270°C for one hour:
(ii) Use an air bath as in Fig. 2-13 or Fig. 2-33.
(iii) Heat in an electric furnace maintained at 265"C.
It is advisable to stir the inass occasionally. In all cases
the crucible is allowed to cool in a desiccator, and, before it is
quite cold, it is transferred to a warm, dry, glass-stoppered
tube or bottle, out of which, when cold, it may be weighed
rapidly as required. It is important to remember tha~
anhydrous sodium carbonate is hygroscopic and exhibits a
tendency to pass into the mOllohydrate. .
Procedure. Weigh out accurately from a weighing bottle
about 0·2 g. of the pure sodium carbonate into a 250 ml.
'conical flask (1), dissolve it in 50 to 75 m!. of water, and add
2 drops of methyl orange indicator (2) or preferably of methyl
orange-indigo carmine indicator (Section I, 29);which gives
a very much more satisfactory end point.* Rinse a clean
burette three times with 5 ml. portions of the acid ;~ fill the
burette to a point 2 to 3 cm. above the zero mark and open"
the stop-cock momentarily in order to fill the jet with liquid.
Examine the jet-to see that no air bubbles are enclosed. If
there are, more liqui~ must 1;>e run out until the jet is com-
pletely filled. Refill, if necessary, to bring the level above the
zero mark; then slowly run out the liquid until the level is
between the 0'0 and 0·5 ml. mark. Read the position of the
meniscus to 0·01 ml. '(Section II, 22), Place the conical flask
containing the sodium carbonate solution uPQn a piece of
'unglazed white paper beneath the burette, and run in the acid
slowly from the burette. Du.ring the addition of the acid, the
flask must be constantly rotated with one hand whilst the
other hand controls the stop-cock. Continue the addition
until the methyl orange becomes a very faint yellow, or the
green colour commences to become paler when the methyl
orange-indigo carmine indicator is used. Wash the 'walls
of the flask down with a little distilled water from a wash
* This indicator is prepared by dissolving I g. of methyl orange and 2·5 g.
of purified indigo carmine in 1 litre of distilled water. and filtering the solution.
The colour change on passiIlg from alkaline to acid solution is from green to
magenta with a neutral grey colour at pH of about 4.
Volumetric Analysis 281t
bottle, and continue the titration very carefully by adding the
acid dropwise until the colour of the methyl orange becomes
orange or a faint pink, or the colour of the mixed indicator is a
neutral srey'; if the end point is overstepped with the latter
indicator, the colour is magenta: This marks '~he end poifit
of the titration, and the burette reading should be taken and
recorded in a note book. The procedure is repeated with two
or three other portions of sodium carbonate. The first (or
preliminary) titration will indicate the locatibn of the true
end point within 0·5 ml. With experience and care, subse-
quent titrations can be carried out very accurately and should
yield concordant results. From the weights of sodium
carbonate and the volumes of hydrochloric acid employed, the
strength of the acid may be computed for each titration. The
arithmetical mean is taken as the strength of the solution.
Notes. 1. For elementary students, an approximately O·IN
solution of sodium- carbonate may be prepared by weighing out
accurately about 1·3"g. of pure sodium carbonate in a weighing bottle
or in a small beaker, transferring it to a 250 ml. volumetric flask,
dissolving it in water (Section n, 6, A (iii)), and making .up to
the mark. The flask is well shaken. 25 Ml. portions are with-
drawri with a pipette and titrated against the acid as described
above. Individual titratiops should not differ by more than 0·1 ml.
2. Alternatively, bromo-phenol blue or methyl yellow may be used
as indicators. With the former a few drops of indicator solution
impart a blue colour to the solution, which becomes purplish-green
at the end point; with the latter the colour changes are ~imilar to
.
those for methyl orange, but are more readily. perceived .
Calculation of normality. The normality may be computed
from the equation:
Na 2CO S + 2HCl = 2NaCl + CO + 1! 0,
2 2
but the best method is to derjve the normality entirely in

terms of the primary standard substance, here, sodium
carbonate. The equivalent weight (Sec'tion I, '23) of sodium
carbonate is 52·997 or 53·00. If the ","eight of the sodium
carbonate is divided by the number of ro!. of hydrochloric
acid to which it is equivalent, as found by titration, we have
the weight of primary standard equivalent to 1 ml. of the
acid. Thus if 0·2500 g. of sodium carbonate is required for
the neutralisation of 45'00 ml. of hydrochloric acid, 1 ml. of
the acid would be equivalent to 0'2500/45'00 = 0·05556 g.
of sc1dium carbonate. The milli-equivalent weight or the
weight in 1 ml. of N -sodium carbonate solution is 0·05300 g.
284 Quantitative Inorgahic Analysis '"
Hence the normality of the acid is 0'05556/0·05300 -
0·1048N.
Another method is the following. 0·2500 G. of sodium
carbonate requires 45'00 m!. of acid, hence 1 litre of acid is
e4_uivalent to 1000 X 0'21)00/45'00 =, 55·56 g. of sodium
carbonate. But a litre of N acid is equivalent to 53·00 g. of
sodium carbonate, hence the acid is 55'56/53·00 = 0·1048N.
In the method described in Note 1 above, the normality of
the sodium carbonate is first computed front the weight of
sodium carbonate used. The mode of calculation described in
Section I. 24 is employed. If VA is the volume in ml. of the
standard solution of normality nA required to react completely
with VA m!. of the unknown solution of normality n then
B,
VA X n A = VB X nil,
from which the value of n A is readily deduced, Thus if
1·3890 g. of anhydrous sodium carbonate is dissolved in 250
ml. of water, the normality of the sodium carbonate solution,
is 1·3890 X 4/53·00 '"= 0·1048N. If 25 m!. of the sodium
carbonate solution exactly neutralise 25·45 ml. of the hydro-
chlonc acid, then
25·00 X 0·1048 = 25·45 X nB ,
or the acid is O· I030N.
B. Standardisation against borax. The advantages of
sodium tetraborate dece.hydrate are: (i) it has a large equiva-
lent weight, 190·72 (that of anhydrous sodium carbonate is
53'00) ;_ (ii) it is easily and econo~ically purified by recrys-
tallisation; (iii) heating to constant weight is not required;
(iv) it is practically non-hygroscopic; and (v) a sharp end
point can be obtained with methyl red at room temperatures
since this indicator has no effect upon the very weak boric
acid.
Na 2 B4 0 7 ,10H 20 +- 2HCl = 2NaCl + 4HaBOa + 5H 20.
Pure borax. A.R. Borax is recrystallised from distilled
water; 50 ml. of water is used for every 15 g. of borax. Care
must be taken that the crystallisation does not take place
above 55°C; above this temperature there is a possibility
. of the formati9n of the pentahydrate since the transition
temperature, decahydrate ~ pentahydrate, is 61°C. The
crystals are filtered at the pump (Section II, 12). washed
twice with water, then twice with portions of 95 per cent
alcohol, followed by two portions of ether. 5 MI. portidn's of
water, alcohol or ether are used for 10· g. of borax crystals.
Volumetric Analysis 285
Each w~shing must be followed by suction to remove the
wash liquid. After washing with alcohol and ether, the borax
is spread in a thin layer on a watch or clock glass and allowed
to stand at room temperature for 1-2 to 18 hours. The borax
is t.hen dry and may be kept in a well-stoppered tube for
3 to 4 weeks without appreciable change. An alternative
method of drying is to place the recrystallised product after
having been washed' twice with water .in a desiccator .over a.
solution saturated with respect to sugar (sucrose) and sodium
chloride. The borax is dry after about'3 days, and maybe
kept indefinitely in the d~siccator without change. The
latter method is more time-consuming; the product is
identical with tha.t obtained by the alcohol-ether process.
Procedure. Weigh out accurately from a weighing bottle
0·4 to 0·5 g. of pure borax into a 250 m!. conical flask (1),
dissolv~ it iq about 50 ml. of water and add a few drops of
methyl red (2). Titrate with the hydrochloric acid contained
in a burette (for details,_ see under A) until the colour changes
to pink (3). Repeat the titration with two other portions of
borax. Calculate the strength bf the hydrochloric acid from
the weight of borax and the volume of acid used: The,
average deviation of these results should not exceed 1 to 2
parts per thousand. If it is greater, further titrations must
be performed until the average deviation is within these
limits. The arithmetical mean is taken as the concentration
of the solution. •
Notes. 1. For elem~ntary students. an approximately O'IN
solution of borax may be prepared by weigbing out accurately 4·7
to 4·8 g. of A.R. borax on a watch g!ass or in a small beaker, trans-
ferring it to a 250 m]. volumetric flask, dissolving it in water (Section
II, 6, A (iii) ), and making up to the mark. The contents of the
flask are well mixed by shaking. 25 Ml. portions--are withdrawn
with a pipette and titrated against the acid as detailed under Method
A. Individual titrations should not differ by more than 0·1 ml.
2: Methyl orange, methyl orange-indigo carmine or bromo-
cresol green may also be used as indicators (compare Note 2 in
Method A). If the comparison solution procedure is adopted (see
Note 3), the mixed indicator should not be used.
3. For work of th~ highest precision,a comparison solution or
colour standard may be prepared for detecting the equivalence
pom. For O·IN solutions; this is made by adding 5 drops of
methyl red to a solution containipg 1,0. g. of sodium chloride and
2·2 g. of boric acid in 500 mi. of water ;_ the solution must.be boiled
to remove any carbon dioxide which may be present in the water.
It is assumed that ~ mI.· of wash water are used in the titration.
286 _Quantitative Inorganic ?f. nalysis
Calculation of the normality. This is carried out as .des-
cribed in Method A. The equivalent weight of borax is
190·72.-
III, 5. Preparation of !;tandard aIkali.-Discussion. The
hydroxides of sodium, potassium and barium are generally
employed for the preparation of solutions of standard alkalis.
They are strong bases and are readily soluble in water.
Solutions made from ammonium hydroxide are undesirable,
because they tend t9' k>se ammonia, especially if the concen-
tration exceeds 0·5N; moreover, it is a weak base and diffi-
culties arise in tit rations with weak acid (compare Section
1, 36). Sodium hydroxide is most commonly used because
of its cheapness. None of these solid hydroxides can be
obtained pure, so that a standard solution cannot be prepared
by dissolving a known weight in a definite volume of water.
Both sodium and potassium hydroxides are extremely hygro-
scopic;' a certain amount of alkali carbonate and water are
always present. Exact results cannot be obtained in the
presence of carbonate with some indicators, and)t is therefore
necess~ry to discu~s methods for the preparation of carboJ_late-
efree alkali solutions. For many purposes the A.R. sodium
hydroxide (which contains 1 to 2 per cent of sodium catbonate)
or the comrrrercially.available sodium hydroxide prepared
from metallic sodium (sodium carbonate content < 1 per
cent) i~ sufficiently pure.
. To prepare carbonate-free sodium hydroXide solution one of
several methods may be used: •
(i). Rinse the sticks of A.R. quality, or those prepared from
metallic sodium, rapidly with water; this removes the car-
bonate from the surface. A solution prepared from the
washed sticks is satisfactory for most purposes.
(ii) If a concentrated solution of sodium hydroxide (equal
weights of solid and water) is prepared, covered and allowed
to stand, the carbonate remains insoluble; the clear super-
natant liquid may be poured or siphoned off, and suitably
diluted. (Potassium ca:rbonate is too soluble in the con-
centrated alkali for this method to be applicable.)
(iii) Carbonate may be removed by the addition of a slight
excess of a concentrated solution of barium chloride or of
barium hydroxide, shaking well, and allowing the precipit~e
to settle. The amount· of Ba.(OH)2,8HzO or of BaClz,2H.O
required'is 10 g. or 8 g. re!ipectively for 100 g. of A.R. sodium
hyd;oxide.
(iv) T~e best method, which yields·a product completely
free from carbonate ions, consists in the electrolysis of a
saturated solution,of A.R. sodium chloride with a mercury
cathode and a platinum anode in the apparatus
shown in Fig. 3-1. About 20 to 30 ml. of
redistilled mercury are placed in a 250 ml.
pear-shaped Pyrex separating funnel, over
which is poured about 100 to 125 mI. of an
almost saturated solution of A.R. sodium
chloride. Two short lengths of platinum wire
are sealed into Pyrex glass tubing, one of these
dips into the mercury (cathode) and the other
into the salt solution (anod~). A little mercury
is 'p~aced in the glass tubes and electrical
contact is made by means of amalgamated
copper wires dipping into the mercury in the
tubes. Electrolysis is carried out using 6 FIG. 3-1.
to 8 volts and 0'& to 1 ampere for several
hours; the funnel is shaken at intervals in order to break up
the amalgam crystals that form on the surface of the mercury.
The weight of the sodium dissolved in the amalgam may be
roughly computed from the total current passed; the current
efficiency is 75 to 80 per cent. When sufficient 'amalgam has
formed, the mercury is run into a Pyrex or, preferably, a
J ena flask containing about 100 ml. of boiled-out distilled water
and closed with a rubber"bung carrying a soda lime guard
tube. Decomposition of the- amalgam is complete after
several days: after 12 to 18 hours about 75 per cent of the
amalgam is decomposed. Dilute solutio~s of sodium hydroxide
are prepared by transferring t~e concentrated solution to
the appropriate quantity of boited-out (or conductivity)
water, an atmosphere of nitrogen being maintained through-
out. This method. is rather time-consuming, and need only
be used for work demanding the highest precision.
AtteIltion must be directed to the.. fact that alkaline
solutions, particularly if concentrated, attack glass. They
may be preserved, if required, in either a paraffined Mttle*
or a bottle in which the inside is coated with a rubber
t:aint, such as "Thermoprene" (compare Section lIt 25) •
• The clean, dry bottle is warmed slightly and enough molten paraffin wax
.is poured in to give a thick coating. The bottle is rolled gently until an even
thick coating is obtained and the paraffin wax is about to solidify. At this
stage the bottle is stood upright so that an extra thick layer is formed at the
bottom.
Furthermore, solutions of the strong bases absorb carbon
dioxide from the air. If such solutions are exposed to the
atmosphere for any appreciable time they become contamin-
ated with carbonate. This may be prevented by the use of a
storage vessel such as is shown in Figs. 2-25, 2-27 and 2-28 ;
the guard tube should be filled with soda-lime or " sofnolite."
A short exposure of an alkali hydroxide solution to the air
will not, however, introduce any serious error. If such
solutions are quickly transferred to a burette and the latter
fitted with a soda-lime guard tube, t.he error due to contamina-
tion by carbon dioxide may be neglected.
The solution of alkali hydroxide prepared by any of the
above methods must be ,standardised. Alkaline solutions
that are subsequently to be used in' the presence of carbon
dioxide or with strong acids are _best standardised against
solutions prepare? from constant boiling point hydrochloric
acid or potassium bi-iodate or sulphamic acid, or against
hydrochloric aCid which has been standar~ised by means of
borax or sodium carbonate. If the alkali sofution is to be
used in the titration of weak acids, it is best standardised
against organic acids, such as benzoic acid or potassi~ hydro-
gen phthalate. The last two are commercially available in a
purity exceeding 99·9 per cent. Potassium hydrogen
phthalate is preferable, since it is more soluble in water and
has a greater molecular weight.
Procedure A. Weigh out rapidly about 4·2 g. of A.R: so-
dium hydroxide on a watch glass or into a small beaker, dis-
solve it in water, make up to 1 litre with boiled-out distilled
water, mix thoroughly by shaking, and pour the resultant
solution into the stock bottle, which should be clo~ed by a
rubber stopper.
Procedure B (carbonate-free sodium hydroxide). Dissolve
50 g. of sodium hydroxide in 50 ml. of distilled water in a
Pyrex or J ena 'las~, transfer to a 75 ml. test tube of J ena or
Pyrex glass, and insert a well-fitting stopper covered with tin
foil. Allow it to stand in a vertical position until the super-
natant liquid is clear. For a O'IN sodium hydrw:ide solution
carefully withdraw 6·5 ml. of the concentrated clear solution
.through a graduated pipette (Fig. 2-18, C) into a litre bottle or
flask, and dilute quickly with 1 litre of recently boiled-out
water.
A clear solution can be obtained more quickly, and inci~
,
dentally the transfer can be made more satisfactorily, by
filtering the solution through a sintered glass funnel with
exclusion of carbon dioxide with the aid of the apparatus
shown in Fig. 3-2. It ,
is advisable to roughly
calibrate the test-tube
in 5 ml. intervals and
to put the graduations
on a thin slip of paper
gummed to the outside NaOH
of the tube.
Procedure C (carbon-
ate-flee sodium hydrox- FIG. 3-2.
ide.) Weigh out about
5 g. of dry sodium hydroxide sticks or pellets on a watch
glass (m the rough balance, transfer to a 500 m!. Pyrex
beaker, 'and dissolve in about 300 ml. of water. Warm
and add slowly a solution (20 to 25 m!.) prepared from 2 to
2·5 g. of solid crystallised barium ,hydroxide or barium chlor-
ide. Allow the barium carbonate to settle, and decant the
clear solution into a litre bottle or flask; dilute to about I
litre with carbon dioxide-free water and shake well. Allow
the sohition to settle again: siphon or filter off the solution
into a clean bottle fitted with a rubber stopper. Alter-
natively, the solution may be stored in an apparatus similar
to that shown in Figs. 2-25, 2-27 or 2-28; a soda-lime guard
tube should be provided.
TIl, 6. Standardisation of the approximately 0 ·1N sodium
hydroxide.-If the solution contains carbonate (Procedure A)
~ethyl orange, methyl orange-indigo carmine, or bromo-
phenol blue must be used in standardisation against hydro-
chloric acid of known normality. Phenolphthalein or indi-
cators with a similar pH range, which are affected by carbon
dioxide, cannot be used at the or-dinary temperature (com-
pare Section I, 27). With carbonate-free sodium hydroxide
(Procedure B) phenolphthalein 6r thymoL blue (Section I, 33)
may be employed, and 'standardisation may be effected
against hydrochloric acid, potassium bi-iodate, potassium
hydrogen phthalate, benzoic acid, or other organic acids
.(Section In, 7).
Procedure A, With standard hydrochloric acid. Place the
O'IN hydrochloric acid, prepared from the copstant boiling
p~int acid or which has been standardised by means of sodium
carbonate or borax, in the burette. Transfer 2? ml. of the
L
sodium hydroxide solution into a 250 ml. conical flask with
the aid of a pipette (1), dilute with a little 'water, add 1 to 2
drQps of methyl orange or 3 to 4 drops of methyl orange-
indigo ca.t;mine indicator, and titrate with the previously
standardised hydrochloric acid. Repeat the titrations until
duplicate determinatfons agree within 0·1 mI. of each other.
Note. 1. If another burette is available, the alkali solution may
be placed in it, and definite :volumes transferred from the burette
to the conical flask for titrations. This metho-d has the advantage
that volumes of the order of 40 mI. of solution can be conveniently
titrated, thus reducing the experimental error; also, back titration
may be effected if the end point is overstepped.
Calculation of the normality. The normality is readily
computed from the simple relationship
VA X n A = VB X n B ,
where VA and n .. refer to the volume and known ,normality
of the acid respectively, VB is the volume of alkali solution
required for the neutralisation,' and nB is its (unknown)
normality.
Procedure B. With potassium hydrogen- phthalate. The
A.R. potassium hydrogen phthalate has a purity of at least
99·9 per cent*; it is almost non-hygroscopic, but, unless a
product of guaranteed purity is purchased, it is advisable to
dry it at 120°C for 2 hours, and allow it to cool in a desiccator.
Weigh out three 0·6 to 0·7 g. portions 01 the salt into 250 ml.
Pyrex or Hysil conical flasks (1), add 70 ml. of boiled out water
to each portion, stopper each flask and shake gently until the
solid has dissolved. Titrate each solution with the sodium
hydroxide solution contained in a burette, using phenQ).-
phthalein or thymol blue as indicator.
Calculation of normality. This is similat to that described
in Section m, 4. The equivalent weight of potassium
hydrogen phthalate is 204·22. The mean deviation should
not exce,ed 0·1 to 0·2 per cent.
HK.CsH"O" + NaOH = NaK.CsH.O" + H"O.
Note. 1. For elementary students, an approximately O·lN
solution is prepared by weighing out accurately about 5·1 g. of the
ordinary A.R. product, dissolving if in water, and making it up to
250 m!. in a volumetric flask. 25 Ml. portions are employed in the
* The U.S. Bureau of Standards supply a product of 99·97 per cent guaran-
teed purity. Products of 99·97 and 99·95 per cent purity can be obtained
from the Eastman Kodak Co" U.S.A., and the Mallinckrodt Chemical Works,
U.S.A., respectively. See Appendix, Section At 9.
V oiumetric Analysis 291
titrations with the sodium hydroxide solution. Individual titrations
shouIa not differ by more than 0·1 ml. The normality is calculated
in a similar manner to that given in Section m. 4; the equivalent
weightis 204·22.
m, 7. Other standard substances for acidimetry and alka-
limetry.-In addition to the standard substances already
detailed for use in standardising acids and alkalis, numerous
others have been proposed. A number of these will be
briefly described.
A. Sodium oxalate (Na 2C20,; equiv. wt. = 134'02). The
A.R. product has a purity of 99,9 per cent. or better. * It is
recommended that Jhe salt be dried for 2 hours at 120°C
before use in order to remove hygroscopic moisture. Upon
strong heating the salt largely decomposes with the formation
of sodium carbonate:
Na 2C2 0, = Na 2CO a + CO.
lf a little carbon dioxide' is lost by the decomposition of the
sodium carbonate, no error is introduced because- the total
alkalinity is unaffected. Some of the salt (0,25" to 0·3 g.
for a O'lN solution) is carefully weighed into a platinum
crucible, which is then covered with tht'!' lid. The crucible
arid lid are heated, while supported in a hole in a sheet of
asbestos, over a small flame for 15 to 30 minutes until the
sodium carbonate formed just begins to fuse. Towards the
end 'Of the operation the lid is moved aside to bring about'the
complete combustion of the carbon ,formed. The resultant·
carbonate is transferred, to a conical flask or beaker, dissolved
in water, and titrated against the acid solution, using methyl
orange or methyl orange-indigo carmine as indicator. The
disadvantages 6f the method are the possibility of mechanical
loss during the evolution of carbon monoxide and its tedious
nature.
B. Mercuric oxide (HgO; equiv. wt. = 108·31). Mercuric
oxide can be readily obtained pure and dry, and has a com-
paratively high equivalent weight. It dissolves in potassium
bromide solution with the production of potassiuII) mercuric
bromide and an equivalent quantity of potassium hydroxide:
HgO + 4KBr + H 20 = K2[HgBrJ + 2KOH.
A O'IN solution of potassium hydroxide is obtained by
dissolving 10·83 g. of A.R. mercuric oxide jn a solution of 200
• A guaranteed product of 99·95 per cent purity is supplied by the U.S.
Bureau of Standards and also by the Mallinckrodt Cheinical Works, U.S.A.
(see Appendix Section A, 9).
292 Quantitative I norgamc Analysis
g. of A.R. potassium bromide in 300 m!. of water, and diluting
to 1 litre. This solution may be titrated against acids with
inter alia phenolphthalein or methyl orange as ihdicators.
C. Benzoic acid (H.C 7 H sOz ; equiv. wt. = 122·12). The
A.R. product has a purity of at least 99·9 per cent. * For
work demanding the highest accuracy, the acid should be
dried before u~e by careful fusion in a platinum crucible at
about 130°C, and then powdered in an agate mortar. Benzoic
acid is sparingly soluble in water (which is a di~advantage)
and must therefore be dissolved in 95 per cent alcohol. The
mode of use is similar to that alr~ady described Jor potassium
hydrogen phthalate (Section m, 6B). For a O'IN solution, of,
say, sodium hydroxide, weigh out accurately 0·4 g. portions
of the acid into a 250 ml. conical flask, add 10 ml. of alcohol,
shake until dissolved, and then titrate 'the solution against
the strong alkali using phenolphthalein as indicator. A
blank test should be made with the same volume of alcohol
and the indicator: deduct, if necessary, the volume of the
'alkali solution consumed in the blank test.
D. Succinic acid (H Z.C 4H 40(; equiv. wt. = 59'045). The
A.R. product or a: pure commercial product should be re-
crystallised from pure acetone and dried in a vacuum desic-
ca'i:or. The purity is checked by means of a melting point
determination (185°-185·5°q. The- acid is fairly soluble in
wafer; phenolphthalein is a suitable ind~cator. I
E. Furoic' acid (H.CsHaOa; equiv. wt. = 112·OB). The

purified commercial acid, m.p. l!H ~-132°C, is crystallised
once from chloroform and dried at 120°C for 2 hours. The
technical product is purified by one recrysta:Ilisation from hot
'Yater with ·the aid of a little animal charcoal (2 g. of animal
charcoal for 20 g. of acid), dried at 120°C for 2 hours, and
then recrystallised from chloroform as above. The res'lltant
solid is pure. The acid is only moderately soluble in cold
water, and is best dissolved in alcohol as detailed under C
for benzoic acid. Phenolphthalein or an indicator of similar
pH. range.,must be used in the titratioR with a strong base.
F. Adipic acid (H z.C sH s0 4 ; equiv. wt. = 73·07). The
commercial acid is crystallised once from hot water with the
addition of a little animal charcoal, dried at 120°C for 2 hours,
then recrystallised from acetone and again dried at 120°C
•
* A guaranteed product of 99·95 per cent. purity is supplied by the U.S.
Bureau of Standards: An acid of similar purity can also be obtained from the
Mallinckrodt Chemical Works, U.S.A. See Appendix, Sections A, 8,9;
V q,lumetric Analysis 293
for 2 hours. The pure acid has a melting point of 152°C.
The acid is slightly soluble in cold water, but, like furoic
q_cid, it dissolves readily in hot water and can be titrated in
hot solution with an indicator of the phenolphthalein type.
It is best, however, to dissolve it in rectified spirit (compare
benzoic acid above). This is an excellent and inexpensive
standard.
G. Potassium. bi-iodate (KH(IO a)2; equiv. wt. = ~89·95).
Unlike the other solid standards already described, this .is a
strong acid and thus permits the use of any indicator having a
pH range between PH 4·5 and 9·5 for titration with strong
bases (compare Section I, 38). It may be employed for the
standardisation of bases which are subsequently to be used
with strong acids; an indicator, such as methyl red, must then
be used. The salt is fairly soluble in water, is anhydrous and
non-hygroscopic, and the aqueous solutions are stable for
long periods; the equivalent weight is high. The. only
disadvantage is that the pure substance is expensive. * A
o'IN solution of potassium bi-iodate- contains 38·995 g.
per litre. •
Preparation of pure potassium bi-iodate. 42 Grams of
A.R. iodie acid are dissolved in 179 ml. of water, and half neutralised
by the addition of potassium hydroxide solution prepared from A.R.
caustic potash. The mixture is heated to boiling, six drops of
concentrated hydrochloric acid added, and, after all has dissolved,
the mixture is cooled. The separated potassium bi-iodate is filtered
on a Jena sintered funnel, and purified by three crystallisations from
hot water. The final product is dried at 100°C for several.hours ;
it is 99·97 per cent pure.
H. Oxalic acid dihydrate (H 2 .C2 0.,2H 20; equiv. wt. =
63'034). The A.R. acid has been widely used as a standard,
but in view of the possible uncertainty of the water content
this is not recommended except for elementary students.
The latter objection may be surmounted if the acid is pre-
served in a desiccator over deliquescent sodium bromide.
I. Potassium. tetroxalate (KHC20(,H2C204,2H20; equiv. wt.
= 84·732). This is suitable only for elementary or ap-
proximate work owingi to the possible uncertainty of t-he
water. content.
• A very pure product, prepared from iodie acid of electrolytic origin, is
made by the G. F. Smith Chemical Co., 867 McKinley Avenue, Columbus,
Ohio, U.S.A. The agents for Great Britain are F. W. Berk and Co. Ltd.,
52-54 Leadenhall Street, London, E.C.3, England.
1. Potassium bi-tartrate (KHC 4 H 4 0 6 ; equiv. wt. = 188'18).
The A.R. product has a purity of 99'9 per cent. The sub-
stance is sparingly soluble in cold water and must be titrateIJ
in hot solution, in which it is more soluble.
Phenolphthalein or thymol blue must be used as indicators
for H, 1 and J.
K. Sulphamic acid (NH~S020H; equiv. = 97·09). A
prodnct of comparatively high purity has recently become
commercially available.* It is prepared by the action fo
fuming sulphuric acid upon urea:
CO(NH2h + H 2S0 4 + S03 = 2NH2S0 20H + CO 2,
It is purified by crystallisation from hot water (see below).

The acid is moderately soluble iIi water (2-1'3 g. and 47·1 g. in
100 g. of water at 20°C and 80°C respectively), appreciably
soluble in methanol and ethanol, slightly soluble in acetone,
and practically insoluble in ether; it is a colourless, crystal-
line, non-hygroscopic solid melting with decofnposition at
205°C. Sulphami~ acid is la:rgely dissociated in aqueous
<solution (it is similar to hydrochloric acid in this respect), so
that any indicator with a pH range of 4 to 9 may be employed
with the O'lN solution; the best indicator for use with
strong bases is bIOmo-thymol blue. The substance may be
employed as a primary standard, and, fof general use, is
probably preferable to the expensive potassium bi-iodate.
The aqueous solution is, however, -not ,stable, as hydrolysis
slowly occurs at room temperature with the .formation of
ammonium bisulphate :
N~2S020H + H 20 = NH 4 HSO,.
Purification of sulphamic acid (G. F. Smith, 1938). Dissolve
125 g. of the commerclal product in ZOO ml. .of water preheated 1:0
70°C. Filter the hot solution three times and discard each time the
material crystallising from solution (alfogether about 25 g.); there
will, of necessity, be a lowering of temperature during t;his process.
Cool the final filtrate in a mixture of ice and salt for 20 minutes;
filter on a sintered glass funnel, and wash successively with a small
quantity of ice water, twice with cold ethyl alcohol, and anhydrous
ether. Dry the recrystallised solid in the air for 1 hour, grind it in
an agate or glass mortar, and preserve in a desiccator. The result-
ant sulphamic add is 99·95 per cent pure. The substance may also
be dried at 105°C; decomposition commences at about l35°C.
• From the Eastman Eodak Co., Roche5ter, N.Y., U.S.A.
m, 8. Standard barium hydroxide (baryta) solution.-This
solution is widely employed, particularly for the titration of
organic acids. Barium carbonate is insoluble, so that a clear
solution is a carbonate-free strong alkali. The. equivalent
weight· of Ba(OH)t,8H2 0 .is 167·75, but a standard solution
cannot be prepared by direct weighing owing to the uncer-
tail1ty of the hydration and the possible presence of carbonate.
To prepare an approximately O'lN solution, dissolve 18 g. of
A.R. crystallised baryta (or 20 g_of the commercial substance)
in about 1 litre of water in a large flask. Stopper the flask
-and shake until f.l.ll the crystals have disappeared. Allow'the
solution to stand.for 2 days or until all the barium carbonate
has completely settled out. Decant or siphon off the' clear
solution into a storage bottle of the type depicted in Figs'.
2-25,2-27 or 2-28. A soda-lime guard tube must be provided
to prevent ingress of carbon dioxide. The solution may be
standardised against standard 0 'IN hydrochloric acid or
potassiu,m hydrogen phthalate; phenolphthalein or thymol
blue is employed as indicator.
ANALYSES INVOLVING THE USE OF STANDA~D

ACID AND ALKALI
m, 9. DeterminatioI} of the Na 2CO a content of washing
soda. Procedure. Weigh out accurately about 3·6 g. of the
washing soda crystals, dissolve in water, and make up to 250
ml. in a graduated flask. Mix thoroughly. Titrate 25 ml.
oj the solution with standarp. hydrochloric acid of approxi-
mately O'lN concentration using methyl orange, or, better,
methyl orange-indigo carmine or bromo-cresol green as
indicator (compare Table X). Two consecutive titrations
should agree within 0·1 ml.
Calculation. The weight of anhydrous sodium carbonate
Na 2 CO a which has reacted with the standard hydrochloric
acid can be readily computed from the equation:
Na 2CO a + 2HCI = 2NaCI + H 2 0 + CO 2
106·01 2 X 35·46
The percentag'e of Na 2C0 3 can then be calculated from the
known weight 'Of washing soda employed.
A simpler and more general procedure is to employ the
" normality" method. A normal solution of sodium car-
bonate contains 53'00 g. of Na 2C0 3 per litre; I ml. of a N
solution of sodium carbonate contains 0·05300 g., or. the milli-
equivalent weight. If the volume of hydrochloric acid be
multiplied by the normality we obtain the number of ml. of N
acid. By multiplying the latter figure by the milli-equivalent
weight of sodium carbonate, the weight of Na 2CO a in the
sample titrated is found: I mI. N HCI = 0'05300 g. of
Na 2CO a. The percentage of NaiCOa can then be -readily
calculated.
An actual example will make this clear.
Weight of weighing bottle + substance = 16·7910 g.
Weight of weighing bottle -+" residual substance = 13'OllO g .
.'. Weight of sample used • = 3·7800 g.
This was dissolved in water and made up to 250 ml.
Titration of 25·0 ml. of the carbonate solution against
0'1060N HC!, with methyl orange-indigo carmine as indicator.
Experiment I R.eading 1 I Reading 2 ' Difference

I I 0·00 m!. I ~6'10 m!. 26·10 m!. (p_reliminary)
2 0;00 m!. 25·90 m!. 25·90 m!.

3 0·00 m!. 25·95 m!. 25·95 m!.
Mean
-25·93 m!.
1 Ml. N HCI == 0·05300 g. Na 2CO a.
25·93 X 0·1060 == 2·74'9 m!. N HCI.
2·749 X 0·05300 == 0·1457 g .. Na 2 CO a in portidn titrated.
Weight of washing soda in portion titrated
=3·7800 X 25'0/250 = 0·3780 g .
.'. Perc~ntage of Na 2CO a = 0·1457 X 100/0·3780 = 38·54%
aItematile method of calculation.
25'0 Ml. of the carbonate solution required 25·93 ml. of
0'1060N HCI;
:. 25·0 X normality of carbonate solution = 25·93 X
0'1060, whence the carbonate solution is 25·93 X 0'1060/25·0
= 0·1099N. But N Na 2CO a contains 106'00/2 = 53·00 g.'
Na 2 CO a per litre,
.:. the given solution contains 0·1099 X 53 ·00 = 5'8271 g.
Na 2CO a per litre, and 250 ml. would contain 5',8271 X
250/1000= 1'4568 g.
Thus percentage of Na 2CO a = 1·4568 X 100/3'7800 =
88·54%.
ro, -10. Determination of the stre~ of glacial acetic acid.
-Weigh a dry, stoppered 50 m!. flask, introduce about 2 g.
of glacial acetic acid and weigh again. Add about 20' m!. of
water and transfer the solution q':lantita1tively to a 250 ml.
graduated flask. Wash the small flask several times with
wafer and add the washings to the ;volumetric flask. Make
up ~o the mark -",ith ,distilled, preferably boiled-out, water.
Shake the flask well to ensure thorough mixing. Titrate 25
m!. portions of the acid with 0 ·IN standard sodium hydroxide,
using phenolphthalein or thymol blue as indicator.
NaOH + H.C zH a0 2 = Na:C 2H a0 2 + H 20.
1 Ml. N NaOH . 0'06005 g. H.C 2H a0 2.
Calculate the percentage of H.C2Ha02 in the sample of glacial
acetic acid.
Note on the determination of the acetic acid content of 'Yinegar. '
Vinegar usually contains 4 to 5 per cent of acetic acid. Weigh out
about 20 g. of vinegar as described' above, and make. up to 100 ml.
in a graduated flask. Rem~)Ve 25,ml. with a t>ipette, dilute with an
equal volume of water, add a few drops of phenolphthalein, al1d
titrate with standard O·IN sodium hydroxide solution. As a result
of the dilution of the vinegar, its natural colour will be so reduced
that it will not interfere with the colour change of the indicator.
Calculate the acetic acid content of the vinegar, and express your
result in g. of acetic acid per 100 ml.
m, 11. Determination of sulphuric acid in the concentrated
acid.-Accur_ately yveigh a small 'glass-stoppered weighing
bottle, and add from a clean, dry 2 m!. pipette or measuring
cylinder 0·7 .. to 0·8 ml. of the concentrated acid. Restopper
the bottle immediately and reweigh accurately.'" Place about
100- ml. of water in a 250 ml. volumetric flask, and insert a
short-necked funnel in the mouth of the flask. Pour the
weighed acid into the funnel~ and, without removing the
weighing bottle, wash it thoroughly both inside and outside
with a stream of water from the wash bottle. Rinse the funnel
thoroughly and remove it. Dilute nearly to the mark, and
after 1 to 2 hours, when the solution has acquired the labora-
tory temperature, make up exactly to the mark. Shake and
mix thoroughly. Titrate 25 m!. portions against standard
o'IN sodium hydroxide, using methyl orange or methyl
orange-indigo carmine as indicator.
• • .. Alternatively, a weight pipette (Fig. 2-23, a) may be used. See, however,
Section In. 12.
L*
From the results falculate the perceruage by weight of
H 2S0 4 in the original concentrated acid.
1 Ml. N NaOH == 0·04904 g. H 2SO,.
m, 12. Determination of sulphuric acid in the' fuming
acid (oleum).-Fuming sulphuric acid may beoconsidered as a
mixture of sulphur trioxide and sulphuric acid, althou&,h it
may contain pyrosulphuric acid.H 2S20 7 ; if rich in sulphur
trioxide, it is solid. The acid must not be exposed. to the
atmosphere during weighing and until it is diluted.with water,
as it is extremely hygroscopic. The Lunge-Rey pipette
(Fig. 2-23, b) is particularly ~uitable for this determination;
the lower part of the pipette is protected by a ground-on test
tube ..
Weigh the complete dry pipette (it may be suspended from
. the balance by means of a thin silver wire). Close tap ,2,
open tap 1, and produce a slight vacuum in the bulb by apply-
ing suction (water pump, etc.), and then close stop-cock 1.
ot
Dip the dry point the pipette into the fuming acid (solid
oleum must first be melted by warming gently), and cau-
tiously open the lower stop-cock. As soon as· the :acid has
entered the widened part of the pipette below the lower bulb,
close tap 2. Replace the test tube 4, and weigh again. The
increase in weight will give the weight of acid taken; this
should be about 1 g. Place about 100 ml. of distilled water
in a beaker and cautiously dip the point of the pipette in
the water; open first tap ~ and then tap 2, and allow the acid
to run into the bea,ker. Wash the pipette thoroughly with
distilled. water. Transfer the contents of the beaker to a 250
ml. volumetric flask, make up to the mark, and- shake well.
Complete the determination as in Section m, 11. .
Calculate the pereentage.of S03 in the sample of oleum.
In, 13. Determination of carbon dioxide in the atmosphere.
'-Discussion. The method to be described is due to Petten-
kofer, as modified by Hesse. The carbon dioxide in a known
volume of air is absorbed by excess of standard baryta solu-
tion, and the excess of the latter is determined by titration
against standard oxalic acid or, better, hydrochloric acid,
using phenolphthalein as indicator.
Procedure. The sample of air is collected in a thick-walled
conkal flask of 500 ml. or 1000 m!. capacity. The flask is
fitted with a 2-holed rubber stopper, the holes being closed by .
glass-rod plugs; a mark i~ made on the fias).< at the point
where the rubber stopper ends. The cap.acity of the flask up
to the point at which the rubber stopper reaches into the neck
is determined by filling with water at a known temperature
to this mark.
Fill the flask with air by aspiration with a rubber aspirator,
or by filling it with distilled water and then pouring the water
out slowly, or by any other convenient method: note the
temperature and pressure of the air in the vessel. Add a five-
fold excess of standard barium hydroxide solution (Section
Ill, 8) by removing both glass plugs from the stopper and
inserting the tip of the baryta splution burette into one of the
holes; remove the burette and replace both .glass plugs.
Shake the vessel gently for about 20 minutes, and allow to.
stand for one hour. Remove both plugs, add a few drops of
phenolphthalein, insert the tip of a burette containing stan-
dard hydrochloric acid (0·02 to O'lN) into one of the holes,
and titrate the excess of barium hydroxide present. Repeat
the determination with another sample of air.
Compute the volume of standard alkali which has reacted
with the carbon dioxide: From this volume and the known
volume of the air less the volume of baryta solution added,
calculate the percentage of COli in the air.
1 Ml. O·lN Ba(:OHh =
1·119 ml. of CO; at N.T.P.
m,14. Determination of a mixture of carbonate and hydrox-

ide. Analysis of commercial caustic soda.-Discussion. Two
methods may be used for' this analysis. In the first method
tbe total alkali (carbonate + hydroxide) is determined
by titration with standard acid, using methyl orange,
methyl orange-indigo carmine, or bromo-phenol blue' as
indicator. The carbonate is precipitated with a slight
excess of barium chloride solution, and, without filtering,*
the solution is titrated against standard acid with thymof
blue or phenolphthalein as indicat<1r. The latter titration
gives the hydroxide content, and by subtracting this from
the first titration, the volume of acid required for the carbo-
nate is obtained.
Na 2C0 3 + BaCl2 = BaCO a (insoluble) + 2NaCI;
BaCl2 + 2NaOH = ;Ba(OHh (soluble) + 2NaCl.
* The strong alkalinity of the soluble sodium carbonate due to 'hydrolysis
is replaced by that (PH 8·6) due to a saturated solution of the sparingly soluble
barium carbonate; hence phenolphthalein or thymol blue can then be
employed as indicators.
300 Quantitative· Inorganic Analysis
The second metl:J.od utilises two indicators. It has been
stated in Section I, 37 that the pH of half neutralised sodium
carbonate, i.e., at the sodium bicarbonate stage, is about 8·3,
but the pH changes comparatively slowly in the neighbour-
hood of the equivalence point (compare Fig.l-ll),consequently
the indicator .colour change with phenolphthalein (PH range
8'3-10'0) or thymol blue (PH range (base) 8'0-9·6) is not too
sharp. This difficulty. may be surmounted by using a com-
parison solution containing sodium bicarbonate of approxi:
mat ely the same concentration as the unknown and the same
volume of indicator. A simpler method is to employ a mixed
indicator (Section I, 29) composed o~ 6 parts of thymol blue
and 1 part of cresol red; this mixture is violet at pH 8'4,
blue at pH 8'3, and ros~ at pH 8'2. With this mixed ind,ica-
tor the mixture has a violet colour in alkaline solution and
changes to blue in the vicinity of the equivalence point;
in making the titration the acid is added slowly until the
solution assumes a rose colour. At this stage all the
hydroxide and half the carbonate (by. conversion into bicarbo-
nate) has been neutralised. Let the volume of standard
acid consumed be v m1~
NaOB + HCI = NaCl + H 20:
Na 2CO a '" + HCI = NaHCO s + NaCl.
Another titration is performed with methyl orange or methyl
orange-indigo carmine as indicator. Let the volume of
acid be V mi.
NaOH + HCI = NaCl + H 2 0:
Na 2 CO a + 2HCI = 2NaCl + H 2 0 + CO 2 ,
Then V - 2(V-v) corresponds to the hydroxide, 2(V-v) to
the carbonate, and V to the total alkali. To obtain satis-
factory results by this method the solution titrated musf be
cold (as near O°C as is practicable), and 19S5 of carbon dioxide
must be prevented as far as possible by keeping the tip of the
burette immersed in the liquid. . .
Procedure A. W.eigh out accurately in a glass-stoppered
weighing bottle about 2·5 g. of c.ommercial sodium hydroxide
(e.g., in flaRe form). Transfer quantitatively to a 500 mL
volumetric flask and make up to the mark. Shake the flask
well. Titrate 25 m!. or 50 m!. of this solution with standard
O·IN hydrochloric acid, using methyl yellow, methyl orange,
or methyl orange-indigo carmine as indicator. 'Carry out
two or three titrations ': these should not differ by more than
Volumetric Analysis
·0·1 ml. This gives the total alkalinity (hydroxide + car-
bonate). Warm another 25 ml. or 50 ml. of the solution to
70°C, and add 10 per cent barium chloride solution slowly
from a burette.or pipette in slight excess, i.e., until no further
precipitate is produced. Cool to room temperature, add a
few drops of phenolphthalein to the solution and titrate very
slowly and with constant stirring with standard O·IN hydro-
chloric acid; the end point is reached when the colour just
changes from pink to colourless. If thymol blue is used as
indicator, the colour change is from blue to yellow. The
arflount of acid used' corresponds to th'y hydroxide present.
This method yields only approximate results because of the
precipitation of basic barium carbonate in the presence of
hydroxide. More accurate results are obtained by cons\der..
ing the above titration as a preliminary one in order to ascer-
tain the approximate hydroxide content, and then carrying
out another titration as follows. Treat 25 or 50 ml. of the
solution with sufficient standard hydrochloric acid to neutralise
most of the hy~roxide, then heat and precipitate as bMore.
Under these conditions, practically pure barium carbonate is
precipitated.
1 Ml. N HCl == 0·0401 g. NaOH:.,
1 Ml. N HCl == 0·05300 g. Na 2C0 3 •
Procedure B. The experimental details for the preparatien

of the initial solution are similar to those given under Pro-
cedure A. Titrate 25 m!. or 50 m!. of the cold solution with
~tandard O·IN hydrochloric acid and methyl orange, tht:thyl
orange-indigo carmine, or bromo-phenol blue as indicator.
Titrate another 25 m!. or 50 ml. .of the cold solution, dHuted
with an equal volume of water, slowly against the standard
acid with phenolphthalein or, better, with the thymol blue-
cresol red mixed indicator; in the latter case, the colour at
the end point is rose. .
(:alculate the result as described in the Discussion above.
m, 15. Determination of a mixtUre of carbonate and

bicarbonate.-The two methods available for this estimation
and modifications of those described in the previous Section
for hydroxide-carbonate mixtures. In the first procedure,
which is particularly valuable when the sample contains
'relatively large amounts of carbonate and small amounts of
bicarbonare, the total alkali is first determined in one portion
302 Quantitative 1 norgamc A natys~s
of the solution by titration against standard O·lN hydro-·
chloric .acid ~ith methyl orange, methyl orange-indigo
carmine, or bromo-phenol blue as indicafor :
Na 2CO a + 2HCI = 2NaCI + H 20 +. CO 2 ;
NaHCO a + HCI = NaCl + H 20 + CO 2 ,
Let this volume correspond to V m!. N HCI. To another
sample, a measured excess of standard O'1N sodium hydroxide
(free from carbonate) over that required to transform the
bicarbonate to carbonate is added.
NaHCO a + NaOH = Na 2CO a + H 20.
A slight excess of 10 per cent barium chloride solution is added
to the hot solution to precipitate the carbonate as b.arium
carbonate, and the excess of sodium hydroxid~ sOlUtion imme-
diately determined without filtering off the precipitate by
titration with the same standard acid; phenolphthalein or
thymol blue is used as indicator. If the volume of excess of
sodillm hydroxide solution add~d be equivalent to v ml. of N
sodium hydroxide and v'm!. N acid corresponds to. the excess
of the~ latter, then v-v' = bicarbonate, and V - (v-v') =
carbonate.
In the second procedure a portion of the cold solutioll is
slowly titrated with standard O·1N hydrochloric acid, using
phenolphthalein or,· better, the thymol blue-cresol red mixed
in.d.icator. This (say, M m!.) corresponds to half the carbonate
(compare Section m, 14) :
Na 2CO a + HCI = NaHCO a -+ NaCl. .
Another sample is then titrated against tile same standard
acid with methyl o(ange, methyl orange-indigo carmine or
bromophenol blue as indicator. The volume of acid used
(say, m ml.) corresponds to carbonate + bicarbonate. Hence
2M = carbonate, and m-2M = bicarbonate.
For practice the student should employ sodium sesquicar-
bonate Na 2 CO a,NaHCO a,2H zO (approx.), or else should use
an artificial mixture prepared from 4 g. of A.R. potassium
carbonate and 1 g. of A.R. sodium bicarbonate. The experi-
mental details are similar to those gi\!,en in Section m, 14.
m, 16. Determination of phosphoric acid in commercial
ortho-phosphoric acid.- Discussion. Phosphoric acid is' a
tribasic acid; the pH at the equivalence poin~ in O'lN solu-
tions corresponding to' the primary, secondary and tertiary
stages of ionisation are approximately 4·6, 9·7 and 12·6
respectively (compare Section I, 37). The pH changes in tile
neighbourhood of the equivalence points are not, however,
very :t:narked (see Fig. 1-10). For the first stage (PH 4·6) we
may employ methyl orange or. methyl orange-indigo carmine
mixture (PH range 3 ·1-4·4) or bromo-cresol- green (PH range
.3·8-~:4), but it is best to use a solution of potassium or
sodium dihydrogen phosphate containing the' same amount of
indicator for comparison. For the secona stage (PH 9'7)
phenolphthalein (PH range 8'3-10'0) is only .approximate,
thymolphthalein (PH range 9'3-10·5) is better, but the best
indicator is a mixture, of phenolphthalein (2 parts) and
a-naphtholphthalein (1 part), which gives a change point
from pale'rose through green to violet at PH 9·6. For the
third stage (PH 12·6) no satisfactory indicator is known. If
titration be carried out to the second stage, and then calcium
chloride solution be added to precipitate the trivalent ion as
formed, titration as a tribasic acid is possible with thymol-
phthalein or the mixed indicator. •
Procedure. Weigh out accurately in a ground-stoppered
weighing bottle ~bout 2 g. of commercial syrupy phosphoric
acid. Transfer quantitatively to a 250 m!. volumetric flask
and make up to the mark. Shake well. Titra1:e 25 ml.
portions with standard carbonate-free sodium hydroxide,
using thymolphthalein or, preferably, a-naphtholphthalein-
phenolphthalein as indicator. Carry out two or three titra-
tions; these should not differ by more than 0·1 mI.
HsPO, + 2NaOH = Na 2 HPO, + 2H 20.
1 Ml. N NaOH = 0,04902 g. HsPO,.
Calculate the percentage of HaPO, in the sample.
m,17. Determination of boric acid and borax.-Discussion.

Boric acid acts as a weak monobasic acid -f/{a = 5·8 X 10-10) ;
it cannot therefore be titrated. accurately with O'IN standard
alkali (compare Section I, 34). However, by the addition of
certain organic. polyhydroxy compounds, such as glycerol,
mannitol, dextrose or invert sugar, it is transformed into a
relatively strong monobasic acid, probably of the type:
1 1
-C-,-OH HO,,- . -C-O"
1 + /B-OH- '1 _ /B -OH + 2H 0.
2
-C-OH HO -C-<Y
1 I
Glycerol is widely employed; the commercial substance is
often acidic, consequently it is necessary to make the dilu-
ted glycerol neutral to phenolphthalein by adding sodium
hydroxide solution. Mannitol (mannite) is more effective and
has the advantage of being a solid and does not materially
increase the volume of the solution being titrated: 0·5 to 0'7
g. of mannitol to 20 ml. of solution is a convenient. quantity.
H [boric acid complex] + NaOH
= Na [bQric acid complex] + H 2 0.
1 ml. N,NaOH == 0·06184 g. HaBOa.
The method may be,. applied to boric acid of commerce and
to borates. The technical boric (l.cid may contain ammonium
salts. It is best to dissolve 1 g. in hot water, add a slight
excess of sodium carbonate, boil the solution down to half
bulk in order to expel ammonia, filter any precipitate which
may be formed, and to titrate the filtrate against standaw
O·IN sodium h:tdroxide with phenolphthalein' as indicator.
For practice in the titrativn the student may employ
commercial borax. This may be titrated~against standard
acid with methyl orange or with methyl orf-nge-indigo
carmine mixed indicator (Section I, 37) :
Na2B407 + 2HC1 + 5H2 0 = 4H 3 B0 3 + 2NaCl.
1 Ml. N HCl =.0,10065 g. Na 2 B 4 0 7 •
The liberated boric acid may then be titrated against standard
alkali in 'the presence of mannitol with phenolphthalein as
indicator. Carbonic acid must be absent as it leads to high
results. -
1 Ml. N NaOH = 0·06184 g. H 3B0 3 == 0·05033 g.'Na 2B 40 7 •
Procednre. Weigh out accurately about 4 g. of commercial
borax, transfer it quantitath::ely to a 250 ml. volumetric
flask, and make up tq the mark. Shake welL. Titrate 25 ml.
of this solulion with staridard O·IN l)ydrochloric acid, using
methyl orange or methyl orange-indigo carmine as indicator.
To another 25 ml. of·the solution, add the quan.tity of
standard hydrochloric acid determined in the previous
titration, * cover the beaker or conical flask with a clock glass
or Fisher "speedyvap" beaker cover, heat to simmering
temperature for 10 minutes to expel carbon dioxid~ (pro-
longed vigorous tfoiling may cause loss of boric acid which is
• If carbonate is absent, mannitol may be added at this stage, and the
titration completed with standard sodium hydroxide sQlution ill the presence
of phenolphthalein.
_ Volumetric Analysis 305
volatile in steam), cool the solution to room temperature, and
add just sufficient standard sodium hydroxide solutio.n. to
restore the transition colour of the original indicator. Then
introduce about 2 g. of mannitol, and shake until dissolved;
add a few drops of phenolphthalein and titrate against
standard O'llY carbonate-free sodium hydroxide to the. first
permanent faint pink {(olour. Add a further 0·5 g. of manni-
tol; if the solution becomes colourless, add more standard
sodium hydroxide until the pink colour reappears. Repeat
the process until the additipn of mannitol has no effect upon
the end point. .
Calculate the percentage of Na zB 40 7 in the sample from the
results of both titrations.
. m, 18.. Determination of ammonia in an ammonium salt:-!-
.
Discussion. Two methods, the direct and indirect, may be
used for this determination. In the direct method, a solution
of the ammonium salt is treated with a solution of a strong
base (e.g., sodium hydroxide) and the mixture distilled.
Ammonia is quantitatively expelled and is absorbed in an
excess of standard acid. The excess of acid is back-titrated
in the presence of methyl red (or methyl orange, methyl
orange-indigo carmine, bromo-phenol blue or bromo-cresol
green). Each ml. of N acid consumed in the reaction is
equivalent to 0·017032 g. NH 3.
NH,Cl + NaOH = NaCl + NH3 + H 20.
In the indirect method, " the
ammonium salt (other than the B
carbonate or bicarbonate) is
D
boiled with a known excess of
standard sodium hydroxide sol-
ution. The boiling is continued
until no more ammonia escapes
with the steam. The excess of
sodium hydroxide is titrated
against standard acid, using
methyl red (or methyl orange-
indigo carmine) as indicator.
Procedure (direct method). Fit
up the apparatus shown in Fig.
3-3. A is.a 1 litre Pyr~x or Jena
E
glass ;flask provided with a two-
holed rubber stopper which
carries a tap funnel B and a FIG. 3-3,
306 Quantitative Inorganic A nqZysis
glass spray trap C. The purpose of the trap' is to prevent
any -droplets of sodium hydroxide solution being driven
over in the subsequent dist,illation. D is a Liebig's condenser
attached to the trap by means of a rubber bung, and E is a
350 ml. Pyrex conicalllask which serves as a receiver. For
practice, weigh out 1·5 g. of A.R. ammonium chloride or
of crude sal ammoniac, dissolve it in water, and make up to
250 mI.-in a volumetric' flask. Shake thoroughly. Transfer
50 mI. of the solution into the distillation flas~ and dilute
with 200 ml. of water; add a few pieces (ca. I gJ of pure zinc
foil Qr granulated zinc to prbmote regular ebullition in the
subsequent distillation. Place 100 mI. of standard O'IN
hydrochloric acid in the receiver and .adjust the fl~k so that
the end of the condenser just dips into the acid. Make sure
that all the corks are fitting tightly. Place 100 mI. of 10 per
cent sodium hydroxide solution in the funnel. Run the
sod~um hydroxide solution into the flask by opening the tap
F; close the tap as soon as all the alkali has entered. Heat
the flask so that the contents boil gently. Continue the dis-
tillation for 30 to 40 minutes by which time all the ammonia
should have passed over into the receiver. Open the tap F
before removing the flame. Disconnect the trap from the
top of the condenser. Lower the .receiver and rins'e the
condenser with a little water. Add a few drops of ,methyl red
and titrate the excess of acid in the solution with standard
O'IN sodium hydroxide. Repeat the determination:
Calcula,te the percentage of NHa in the solid ammonium
salt employed.
Procedure (indirect method). Weigh out accurately 0·1 to
0·2 g. of the ammonium salt into a 350 ml. Pyrex or J ena glass
conical flask, and add 100 mI. of standard o'IN sodium hy-
droxide. Place a small funnel in the neck of the flask in order
to prevent mechanica11oss, and boil the mixture until a piece
of filter paper moistened with mercurous nitrate solution is
no longer turned bla~k. Cool the solution, add a few drops
of methyl red, and titrfl.te with standard O·IN hydrochloric
acid. Repeat the determination.
Calculate the percentage of NHa in the substance.
m, 19. Determination of nitrates 1sodium nitrate in Chile
saltpetre).-Discussion. Nitrates are reduced to arn.monia by
means of aluminium, zinc or, most conveniently, by Devarda's
alloy (50Cu, 5Al, 5Zn) in strongly alkaline solution:
3NO a- + SAl + 50H- + 2HzO =- SAI02- + 3NH s•
The ammonia is distilled into excess of standard acid as in
Section m, 18.
Nitrites are similarly reduced and must be allowed for if
nitrate alone is to b~ determined. .
Procedure. Weigh out accurately about I g. of Chile salt-
petre. Dissolve it in water and transfer the solution quanti-
tatively to the distillation flask of Fig. 3-3. Dilute to about
240 ml. Add. 3 g. of pure finely-divided Devarda's alloy (it
should all pass a 20-mesh sieve). Fit up the a.Rparatus
completely (all connexions must be made with rubber bungs),
and place 75-100 ml. of standard 0'2N hydrochloric acid in
th~ receiver (500 ml. Pyrex conical flask). Introduce 10 ml.
of 20 per cent sodIum hydroxide solution through the funnel
and immediately close' the tap. Warm gently to start the
reaction, ~nd allow th.e apparatus to stand fur an hour, by
which time the evoluti.on of hydrogen should have practically
ceased and the reduction of nitrate to ammonia be complete.
Then boil the liquid gently and continue the distillation ul!til
40-50 ml. of liquid remain in the distillation flask. Open the
tap before removing the flame. Wash the condenser with a
little distilled water, and titrate the contents of the receiver
and the washings against standard 0'2N sodium hydroxide,
using methyl red as indicator. Repeat the determination.
For very accurate work, it is recommended that a blank test
be carried out with diStilled water.
Calculate the percentage of NaNO a in Chile saltpetre.
1 Ml. N HCl lSi 0·08501 g. NaNO a•
Note. Nitrates can be reduced in neutral solution quantitatively
to ammonia with Arndt's alloy (60 per cent Mg; 40 per cent eu) ;
any slight spurting' of the boiling liquid does not vitiate the titra-
tion. Here the distillation· flasks will contain 50 ml. of the above
nitrate solution, 150 ml. of water, 2 g. of hydrated magnesium chlor-
ide, and 5 g. of the finely-divided alloy. The wIlole is boiled until
about two thirds of the liquid has passed over. The distillate is
collected in 50 mI. of O'lN hydrochloric acid.
m, 20. Determination of nitrogen by Kjeldahl's method.-
D~scussion. T-his process is so widely employed in industrial
and research laboratories that an account of it must be
included in this book. The principle of the met!lod is the
conversion ot the nitrogen of the nitrogenous substance by
boiling with concentrated sulphuric acid into ammonia, which
is fixed by the excess of acid as ammonium su1phate. The
latter is determined by adding an excess of caustic alkali to
the solution after digestior with the acid, and distilling off
the liberated aItlmonia into standard acid as detailed in Sec-
tion III, 18. Simple digestion with concentrated sulphuric
aciJ -is, however, a slow process and various modifications
have been suggested to increase the st>eed of the reaction.
These include the addition of potassium sulphate, which
raisei the boiling point of the acid (Kjeldahl-Gunning
process), and 'of catalysts, such as' mercury, mercuric oxide,
anli'ydrous copper sulphate, selenium, or a mixture of equal
parts of ferrous sulphate and selenium. The simple process
works well for nitrogen determinations in proteins and also
in amines and amides, but is not applicable to nitro, azo ..
hydrazo, and cyano compounds without modification. In
the presence of nitrates, there is a danger of loss of nitric
acid: salicylic acid is then added to the sulphuric acid
(proportion: :t g. to ,30. mI.), which fixes the nitric acid as
nitro-salicylic acid. Upon the addition of zinc dust (1 g.) or
of anhydrous sodium thiosulphate (5 g.), the nitro-s~licylic
acjd is reduced to the amino compo}lnd, which can then be
estimated by the Kjeld'ahl process. -
Procedure. The determination may be carried out with
1·5 to 2·0 g. of flour or with acetanilide. Weigh out aCC)lr-
ately about 0·1 g. of acetanilide into a dry Pyrex Kjeldahl
flask-a round-bottomed flask with a long narrow neck. Add
10 g. of A.R. potassium sulphate, 0·2 g. of sel~niuljll or 0·5 g. of
anhydrous copper sulphate or 0·5 g. of equal parts of Jerrous
sulphate and selenium, and pour 25 mI. of concentrated
sulphuric acid into the flask in such a 'way as to wash down any
solid adhering to the neck. Shake until the contents are well
mi:l(:ed. Support the flask on a hole (4·5 cm. in diameter)
in a piece of asbesto's board, and incline the neck at, an angle
of about 60°. Close the flask with a loosely-fitting glass
stopper elongated to a point and having a balloon-shaped top:
a small. funnel or a flat reagent bottle stopper will serve the
same purpose. Heat the flask in a fume cupboard with a
small flame. When frothing has ceased, increase the flame
until the mixture boils gently. Continue the heating for 30
minutes after the solution has become cold'urless or' clear
(usually 90 to 120 minutes)" and then allow Hie solution tq cool.
When cold, add about 200 ml. of water carefully and with
frequent shaking. Transfer completely to the ammonia
distillation apparatus (Fig. 3-3), add a little pure zinc, followed
by excess of sodium hydroxide solution. Distil off tre
ammonia into excess of standard acid exactly as described'in
Section lIT, 18. Carry out a blank determination exactly as
Volumetric A nalrsis 309
above, but with the nitrogen-containing sample omitted
subtract the" blank," if any.
C~lculate the percentage of nitrogen in the solid.
Note~ If mt!rcury (0·7 g.) or mercuric oxide (1 g.) is employed as

catalyst, it is necessary to add 25 ml. of a solutiod containing 8 g.
of sodium thiosulpliate or 4 g. of potassium sulphide per 100 ml.
before distilling in order tl? decompose the mercuri-ammonium
compounds formed.
PRECIPITATION AND COMPLEX-FORMATION
REACTIONS
m, 21. General discussion.-The reader should be familiar
with the theory of precipitation and complex .formation
processes given in Sections I, 40 to I, 42. W.ith this theoretical
groundwork, he should experience little difficulty in carrying
out the following determinations in an intelligent manner,
and also to modify them when occasion demands.
The stand~rd solutions employed include those of silver
nitrate, sodium chloride; potassium or ammonium thiocya-
nate, and potassium cyanide.
m, 22. Preparation of O'IN silver nitrate.-Discussion.
Very 'pure silver can be obtained commercially, and a stan-
dard ~olution can be prepared by dissolving a known weight
(say, 10·788 g.) in pure- dilute nitric acid in a conical flask
having a funnel in the neck to prevent mechanical loss, and
making up to a known volume (say, 1 litre for a O·lN solution).
The presence of acid .....must, however, be avoided in such
determinations as that of Mohr (Section I, 42A) with potas-
sium chromate as indicator or in estimations e:Q1ploying
adsorption indicators (Section I, 42C). -It is therefore pre-
ferable to employ a neutral solution by dissolving silver
nitrate (equivalent weight, 169·~m) in water. I _
A.R. Silver nitrate has a purity of at least 99·9 per cent, so.
that a standard solution can be prepared by direct weighing.
If, .however, the ordinary recrystallised silver nitrate be
employed, or if an additional check of the normality of the
silver nitrate solution is required, standardisation may be
effected with pure sodium chloride:
AgN0 3 + NaCI = AgCl + NaNO a•
The ordin'ary A.R. sodium chloride has a purity of 99·9 to
100·0 per cent*; the substance is therefore an excellent primary
standard. Sodium chloride is verY' slightly hygroscopic,
and for accurate work it is best to dry the finely-powdered
solid. in an electric oven at 250° to 350°C for 1 to 2 hours, and
allow it to cool in a desiccator.t
... Sodium chloride is supplied by the Mallinckrodt Chemical Works. U.S.A .•
of 99·97 per cent purity, and by the Bureau of Analysed Samples, as a British
Chemical Stap.dard. of 99·96 per cent purity. See Appendix. Section A. 9.
t For the most precise work. the powdered solid should be dried in an electric
muffle furnace at 500° to 600°C and then allowed to'cool in a desiccator. This
high temperature is unnecessary for work of ordinary accuracy (within 0·1
per cent).
310
Procedure ·A. From A.R. silver qitrate. Dry some finely-
powdered A.R. silver nitrate at 150°C for 2 hours and allow
it to cool in a desiccator. Weigh out accurately 8·496 g .• ,
dissolve it in water and make up to 500 ml. in a volumetric
flask. This gives an 0·1000N solution. Alternatively, about
8·5 g. of the p~re, dry silver nitrate may be weighed out
accurately, dissolved in 500 mI. of water in a graduated flask,
and the exact normality computed from the weight of silver
nitrate employed.
Procedure 13. From recrysta1lised silver nitrate. Weigh out
about 8·5 g. of the silver nitrate, dissolve it in water, and make
up to 500 mI. in a volumetric flask. Shake well.
Solutions of silver nitrate should be 'protected from light;
they are best stored in amber-coloured &_lass bottles.
m, 23. Standardisation of the silver nitrate solution.-
Sodium chloride ha$ an equivalent weight of ,58·~6. A
0'1000N solution is prepared by weighing out 2·923 g. of the
pure dry salt (e.g., A.R., but see Discussion in Section In,
22) and dissolving it in 500 mI. of water in a volumetric flask.
Alternatively about 2,9 g. of the pure salt is accurately
weighed out, russolved in 500 mI. of water in a graduated
flask, and the exact normality calculated from the weight of
sodium chloride employed.
A. With potassium. chromate as indicator. The.Mohr titration.
The student is referred to Section I, 42B for the detailed
theory of the titration. Prepare the indicator solution
by dissolving 5 g. of A.R. potassium chromate in 100 ml.
of water. The fina1 volume of the solutio:{l in the titration is
50 to 100 mI., and 1 m!. of the indicator solution is used, so
that the indicator concentration in the- actual titration is
0,005 to 0·0025 molar.'
Pipette 25 mI. of the standard 0'1N sodium chloride into a
250 ml. conical flask resting upon a white tile (1), and add 1
mI. of the indicator solution (preferably with a 1 mI. pipette).
Add the silver nitrate solution slowly from a burette, swirling
the liquid co.nstantly, until the red colour formed by the
addition of each drop begins to disappear more slowly: this
is an indication that most of the chloride has be~n precipi-
tated. Continue the addition dropwise until a faint but
distinct change in colour occurs. This faint reddish-brown
colour should persist after brisk shaking. If the end point is
* This weight should- be multiplied by the purity factor of the A.R. silver
nitrate, e.g., by 1/0'999.
•
312 Quantitqtive Inorganic Analysis
overstepped (production !?f a deep reddish-brown colour),
add more of the chlor-ide solution and titrate again. Deter-
mine the indicator blank correction by adding 1 ml. of the
indicator to a volume of water equal to the final volume in the
titration (2), and then silver nitrate solution until the colour
of the blank matches that of the solution titrated. The
indicator blank correction, whi~h should not amount to more
than 0·03 to 0'10 ml. silver nitrate, is dedUcted from the
vplume of silver nitrate used. in the titration. Repeat the
titration with two further 25 ml. portionS" of the sodium
chloride solution (3). The various titrations s40uld agree
within 0·1 ml.
Calculate the normality of the silver nitrate solution (see
Section m, 4A). " .
Notes. 1. The end point • is very readily detected in a large
porcelain basin or in a casserole. The solution is stirred with a
short glass stirring .rod.
2. A better blank is obtained by adding about 0·5 g. of A.R
calcium carbonate before determining the correction. This give
an inert white precipitate similar to that obtained in the titration
of chlorides and materially assistS" in matching th~ colour tints.
3. See Note 1 to Procedure A in Section ITl,6.
B. With an adsorption indicator. Discussion.-The de-
tailed theory of the process is given in Section II, 42C. Both
fluorescein and dichlorofluorescein are suitable for the titration
of chlorides. In both cases the end point is reached when the
white precipitate in the greenish-yellow solution suddenly
assumes a pronounced reddish tint. The change is reversible
upon the addition 9f chloride. With fluo~escein the solution
must be neutral or only faintly acidic with acetic acid; acid
solutions should be treated with a slight excess of sodium
acetate. The chloride solutIon shouM be diluted to about
0·01 to 0·05N, for if it is more concentrated the precipitate
coagulates too soon and interferes. Fluorescein cannot be
used in solutions more dilute than 0·005N. With more dilute -
solutions resort must be made to dichlorofluorescein, which
possesses other advantages over fluorescein. Dichlorofluores-
cein gives good results in very dilute solutions (e.g., for drink-
ing water) and is applicable in the presence of acetic acid and
jn weakly acid solutions. For this reason the chlorides of
copper, nickel, manganese, zinc, aluminium and magnesium,
which cannot be t~trated according to the method of Mohr,
can be determined by a direct titration when dichlorofluores-
cein is used as indicator.
For the reverse titration (chloride into- silver nitrate), tar-
trazine (4 drops of a 0·5 per cent solution per 100 ml.) is a
good indicator. At the end 'point, the almost col'ourless
liquid assumes a green colour.
The indicator solutions are prepared as follows:
Fluorescein. Dissolve-O·l g. of fluorescein in 100 m!. of 70 per
cent alcohol, or dissolve 0·1 g. of sodium .auoresceinate in 100 ml. of
water.
Dichlorofiuorescein. Dissolve 0·1 g. of dichlorofluorescein in 100
ml. of 60 to 70 per cent alcohol, or dissolve 0·1 g. of sodium dichloro-
fluoresceinate in 100 ml. of water.
Procedure. Pipette 25 ml. of the standard O'lN sodium
chloride into ~ 250 ml. conical. flask. Add 10 drops :of either
fluorescein or dichlorofluorescein indicator, and titrate against
the silver nitrate solution with constant rotation in a diffuse
light. As th~ end point is approached, the silver chloride
coagulates appreciably and the local development of a pink
colour upon the addition of a drop of the silver nitrate solution
becomes more and more pronounced. Continue the addition
of the silver nitrate solution until the precipitate suddenly
assumes a pronounced pink or red colour. Repeat the
titration with two other 25 ml. portions of the chloride solution.
Individual titrations should agree within 0·1 ml.
. Calculate the normality of the silver nitrate solution.
m. 24. Determination of chlorides.-Either the Mohr
titration or the adsorption indicator. me'thod may be used fot
the estimation of chlorides in neutral solution by titration
with standard O·IN silver nitrate. If the solution is acid,
neutralisation may be effected with chloride-free calcium
carbonate, sodium bicarbonate or borax; the A.R. substances
are suitable. Mineral acid may also be removed b~ neutralis-
ing most of the acid with ammonia solution and then adding
an excess of A.R. ammonium acetate. Titration of the
neutral solution prepared with calcium carbonate by the
adsorption indicator method is rendered more facil~ by the
addition of 5 ml. of 2 per cen't dextrin solution; this offsets
the coagulatin.g effect of the calci~m ion. If the solution is
basic, it 1Jlay be neutralised With chloride-free nitric acid,
using phenolphthalein as indicator.
m, 25. Determination of bromides.-Discussio1t. The
Mohr titration may be applied to the estimation of bro:mides.
The most suitable adsorption indicator is eosin, although
fluorescein may b~ used but is subject to the limitations
detailed under Chloddes in Section ill, 23B. Eosin is service-
able in dilute solutions and even in the presence of O'IN
nitric acid, but in general, however, acetic acid solutions ate,
to be preferred.
The eosin indicator is prepared by dissolving 0·1 g. of eosin in 100
ml. of 70 per cent alcohol, or by dissolving 0·1 g. of the sodium salt
in 100 ml. of water. .-
For the reverse titration lbromide into silver nitrate)"
rhodamine 6G (10 drops of a 0·1 per cent solution of the
chloride) is an excellent indicator. The solution is best
rendered O·5N with respect to nitric acid, and should be 0·01
to O'02N with respect to silver. The precipitate acquires a
violet colour at the end point .•
Another method for the volumetric determination .of bromides,
which utilises an adsorption indicator (bromo-phenol
blue), involves titration with mercurous nitrate solution.
The solid reagent is deliquescent and cannot therefore
be obtained in a state of appropriate purity for accurate
weighing: the A.R. product HgNOs,H 2 0 has a purity of. not
less than 95 per cent. 30 Grams of A.R. mercurous nitrate
.are dissolved in water containing a little nitric acid, and made
up to 1 litre in a volumetrie flask. The reagent is first stan-
dardised against O·lN potassium bromide, prf!pared by dis-
solving 11·902 g. of A.R. potassium bromide in 1 litre of water.
~5 Ml. of the bromide solution is diluted with an equal volume
of water and treated ~ith a few drops of a 0·1 per cent solution
of bromo-phenol blue. Titration is then carried out with the
mercurous nitrate solution until the yellowish tint on the
precipitate suddenly changes to lilac. A similar titration is
then carried out with the unknown bromide -solution.
For practice in the use of eosin, the student may determine
the purity of potassium bromide, preferably of A.R. quality.
Procedure~ Weigh but accurately about 3 g. of the potas-
sium bromide, dissolve it in water, and make up to 250 ml.
in a 'graduated flask. Pipette 2.5 m!. of the well-mixed solution
into a 250 ml. conical flask, add 75 ml. of water, a few ml. of
6N-acetic acid and 10 drops of eosin solution. Titrate
against the standard O'lN silver nitrate, with constant
agitation, in a diffuse light. The silver bromide flocculates
approximately 1 per cent before the equivalence point, and
the lodd development of req colour becomes more and more
pronounced upon the addition of silver nitrate solution.
Continue the addition -dropwise with vigorous swirling of the
V olmnet11c Analysis 315
liquid until the precipitate suddenly assumes a pronounced
magenta colour. Repeat the titration with two other 25 m!.
portions of the bromide solution.
Calculate the purity of the sample .
.
1 Ml. N AgNO a ~ 0·07992 g. Br == 0·1190 g. KBr.
.
m, 26. Determination of chlorides and bromides with
standard -mercurous perchlorate solution. *-Discussion. Mer-
curous perchlorate, with bromo-phenol blue as adsorption
indicator, .may be used for the determination of chlorides and
bromides. It can be applied in acid solution; its solutions
are reasonably stable when kept in the dark (the chloride titre
is reduced by only 0·1 per cent after 4 months), they are easy
to prepare, are free from interfering mercuric ions (compare
mercurous nitrate solution, Section m, 25), and give· sharp
end points which coincide almost exactly with the equivalence
points: The colour change at the end point is from creamy
white to lilac; the best results aTe obtained if the addition
of the indicator is delayed until near the end point.
Nitric, perchloric and acetic acids have no effect if the total
acid concentration does not exceed 0·04N; higher concentra-
tions of acid inhibit the colour change. Sulphuric acid and
sulphates cause a high titre; this error is completely elimin-
ated by adding an excess of lead nitrate to the boiling solution;
the lead SUlphate need not be filtered off before titration. The
only cations which interfere are those of silver and mercuric
mercury; the latter reduces the titre by an amount propor-
tional to their concentration, due presumably to the removal
of some of the halide ions as undissociated mercuric halide.
Procedure. A. Preparation of approximately O'lN mercuro\1$
perchlorate. Shake the following materials vigorously to-
gether at intervals over a period of 1 hour: 22 g. of ~.R. red
mercuric oxide, 45 g. of perchloric acid (d 1,5; 60 per cent),
50 m!. of water, and 30 g. of pure mercury. From time to
time withdraw a few drops of the liquid, dilute, treat with
sodium chloride solution, and filter: the preparation is
complete when the filtrate from the test portion gives no
precipitate with hydrogen sulphide. Decant or filter the
liquid from the excess of mercury, and dilute to 2 litres.
Keep the solution in the dark; if it is exposed to' light,
decomposition into mercury and .mercuric perchlorate occurs.
• W. Pugh. 1937.
316 Quantitative 1no,ganic Analysis
B. Standardisation. Deci-normal sodium chloride (Section
III, 23) or potassium bromide (Section m, 25) may be' em-
ployed. 25 Ml. of either solution is diluted to 100 m!. in a
conical flas~, 0·5 m!. of bromo-phenol blue (0·05 per cent
solution in water) added, and the merturous perchlorate
solution run in from a burette until the colour of the precipi-
tate changes from creamy white to lilaG. It is best to add tlie
indicator near the end point. •
C. Use in determination of chlorides and bromides. Dilute
the solution of the ha,lide to 100 ml. If sulphates are present,
boil, add 10 to 20 ml. of 5 per c~nt lead nitrate -sohltion, and'
cool. Add 1 drop of bromo-phenol blue, then ammonia or
sodium carbonate solution until the indicator turns faintly
blue, and titrate with the O·lN (or 0'05N) mercurous per-
chlorate. Just before the end point is attained, add 0·5 m!.
of bromo-phenol blue, and continue the titration until the
colour of the precipitate changes to lilac.
For practice, the student may determine the percentage of
chlorine in A.R. crystallised barium chloride.
m, 27. Determination of iodides.- Discussion. The Mohr
method cannot be applied to the titration of iodides (and also
of thiocyanates), because of adsorption phenomena an'd the
difficulty of distinguishing the colour change of the potassium
chromate. Eosin is a suitable adsorption indicator, but di-
iododimethylfluorescein is better. Eosin is employed as des-
cribed under bromides (Section m, 25).
The di-iododimethylfluorescein indicator is prepared by dissolving
1·0 g. in 100 ml. of 70 per cent alcohol. The colour change is from
an orange~red to a blue-red on the precipitate.
For p(actice, the student may determine the purity of
potassium iodide, preferably of A.R. quality.
Procedure. Weigh out accurately about 4 g. of the potas-
sium iodide, dissolve it in water and make up to 25<rml. in a
volumetric flask. Pipette 25 m!. of the well-mixed solution
into a 250 m!. conical flask, add 75 m!. of water and 1 mI. of
di-iododimethylfluorescein indicator. Titrate against the
standard O'lN silver nitrate as detailed under Chlorides
(Section m, 23B). Flocculation of the silver iodide occurs
about 1 per cent before the end point, and the colour change
at the end point is from orange-red to blue-red. Repeat
with two other 25 m!. portions of the solution.
Calculate the purity of the sample.
1 M!. N 'AgN03 == 0:1269 g. I == 0·1660 g. KI.
Volumetrio Analysis 317 ,
m, 28. Determination of thiocyanates.-These may be
estimated with adsorption indicators exactly as for Chlorides
and. Bromides. Further experimental details are therefore
superfluous. For another method, see Section m, 43.
m, 29. Determination of mixtures of halides with adsorption
indicators.-A. Chloride and iodide in admixtures. These two ions
differ considerably in the ease With which they are adsorbed on
the corresponding silver halide. This makes it possible to ~elect
adsorption indicators which will permit of the estimation of
chloride and iodide in the presence of one another. Thus the
iodide may be· 'determined with di-iododimefhylfluorescein
(Section m, 2,() by titration with standard O-lN silver nitrate,
and the iodide + yhloride by a similar titration with fluores-
cein. Chloride is obtained by difference. If a large excess
of chloride is present the result for iodide may be as much as
1 per cent high. If, however, Rose Bengal (dichlorotetra-
iodofluorescein) is used as indicator (colour change, carmine
red to blue-red) in the pre~~nce of ammonium carbonate, the
iodide titration is exact.
B. Bromide and iodide in admixture. The total ~alide
(bromide + iodide) is determined by titration with standard
O-lN silver nitrate with eosin or fluorescein as indicator_ The
iodide is estimated by titration with 0-01 to 0-2N silver
nitrate, using di-iodidimethylfiuorescein as indicator. Bro-
mide is obtained by difference.
Numerous adsorption indicators have been suggested for
various purposes, but a full treatment is outside the scope of
this work.*
m, 30. Determination of mixtures of halides by an indirect
method.-Discussion. The method is applicable to the
determination of a mixture of two salts having the same anion
(e.g.) sodium and potassium chlorides) or the same cation
(e.g., potassium chloride and potassium bromiq.e). Let us
first suppose that it ~s desired to estimate the amount of
sodium and potassium chlorides in a mixture of the two salts.
A known weight (WI g.) of the solid mixture is taken, and the
total chloride is determined with standard O'lN silver nitrate,
using Mohr's method or an adsorption indicator. Let w2 g.
of silver nitrate be required for the complete precipitation of
... An excellent account is given by K. Fajans in the Chapter entitled" Ad-
sorption indicators for precipitation reactions" in W. Bottger-R. E. Oesper
Newer Methods of Volumetric Chemical Analysis, 1938, pp. '203-250 (Van
Nostrand; Chapman and Hall).
WIg. of tq.e mixture, which contains x g.- of NaCl and y g. of
Kel. Then:
x + Y = WI;
169·89 x + 169·89 Y _ W
58'46 74'56 - 2'
Upon solving these two simultaneous equations, the values

for x and yare deduced. #
Let us now suppose that the ~stimation of potassium

chloride. and potassium bromide in admixture is desired.
The total halide is determined by Mohr's method or with an
adsorption indicator. Let the weight of th~ mixture be Ws g.,
W, g. be the weight of silver nitrate required for complete
precipitation, p g. be the weight of the potassium chloride,
ILnd q g. be the weight of the potassium bromide. Then:
p + q = wa;
169·98 p 169·89 q
74.56 + 119.;02 = w4 •
The .,alues of p and q can be obtained by solving the simul-
taneous equations. •
~t can be shown* that the method depends upon the differ-
ence between the equivalent weights of the tw<t. components
of the mixture and that inter alia it is most satisfactory when
the two constituents are present in approximately, equal
proportions.
For practice, the student may prepare for himself, or obtain
from the teacher, a mixture 6f equal weights of potassium
and sodium chlorides or of potassium chloride and bromide,
and determine the proportions of each by titration with
standard O'IN silver nitrate. The experimental details for
the tit rations ~ave been given in previous Sections.
m. 31. Preparation and use ot- 0'1N ammonium or potas-
sium thiOCyanate. Titrations according to Volhard's·method.
-Discussion. Volhard's original method (1874, 1878) for
the determination of silver in dilute nitric acid solution by
titration with standard thiocyanate solution in the presence
of a ferric salt as indicator has proved of great value not only
for silver estimations, but also in numerous indirect analyses.
The theory of the Volhard process has been given in Section
I. 42B. In this connexion it must be pointed out that the
* See H. A. Fales Inorganic Quantitative Analysis, 1925, p. 204 (Century),
• Volumetric Analysis 319
concentration of the nitric acid should be from 1 to 10 per
cent by volume (strong nitric acid retards the formation of
the" ferric thiocyanate") and at a temperature not exceed-
ing 25°C (higher temperatures tend to bleach the colour of the
indicator). The solutions must be free from nitrous acid,
which gives a red colour with thiocyanic acid, and may be
mistaken for" ferric thiocyanate." Pure nitric acid is pre-
pared by diluting the usual pure (e.g., A.R.) acid with about
one-fourth of its volume of water and boiling until perfectly'
colourle$s; this eliminates any lower oxides of nitrogen which
may be present.
The method may be applied'.to those anions (e.g., chloride,
bromide and iodide) which are completely precipitated by
silver and are sparingly soluble in dilute nitric acid. Excess
of standard silver nitrate solution is added to the solution
containing free nitric acid,' and the residual silver nitrate
solution'is titrated with standard thiocyanate solution. This
is sometimes termed the residual process. Anions whose
silver salts are slightly. soluble in water, but which are soluble
in' nitric acid, such as phosphate, arsenate, chromate, sulphide
and oxalate, may be precipitated in n.eutral solution with an --
excess of standard silver nitrate solution. The precipitate
is filtered off, thoroughly washed, dissolved in dilute nitric
acid and the silver titrated with thiocyanate solution. Alter-
natively, the residual·silver nitrate in the fiftrate from the
precipitation may b~ determined with thiocyanate solution
after I;lcidifiCfl.tion with dilute nitric acid.
Both ammonium and potassium thiocyanates are usually
available as deliquescent solids; the A.R. products are,
however, free from chlorides and other interfering substances.
An approximately O'lN is therefore first prepared and this is
standardised by titration against standard O'IN silver nitrate.
Procedure. Weigh out about 9 g. of A.R. ammonium thio-
cyanate, or 11 g. of A.R.. potassium thiocyanate, and dissolve
it in I litre of water in a volumetric flask. Shake well.
Standardisation. Use O·IN silver nitrate, which has been
prepared and standardised as described in Section m, 23.
The indicator solution consists of a cold saturated solution of
A.R. ferric ammonium sulphate in water (about 40 per cent) to
wbich a little 6N nitriG acid has been added. 1 Ml. of this solution
is employed for each titration.
Pipette 25 ml. of the standard O·Il\! silver nitrate into a
250 ml. conical flask, add Q ml. of 6N nitric acid and I ml.
of the indicator solution. Run in the potassium or ammonium
thiocyanate solution from a burette. At first a white precipi-
tate is produced, rendering the liquid ·of a milky appearance,
and as' each drop of thiocyanate faIrs in, it produces a reddish-
brown cloud, which quickly disappears on shaking. As
the end P9int approaches, the precipitate becomes flocculent
and settles easily; finally one drop of the thiocyanate 'solu-
tion produces a f~int brown colour, which no longer disappears
upon shaking. This is the end point. The indicator blank
amounts to 0·01 m!. of O·lN silver nitrate. It is essential to
shake vigorously during the titration. in order to obtain
correct results. Repeat the titration with two other 25 m!.
portions of the silver nitrate solution. Individual titrations
should agree within 0·1 ml.
Calculate the normality of the thiocyanate solution from
the known normality of the silver nitrate solution (for details
of calculation, see Sections I, 24 and m, 4A). It will be found
that the thiocyanate solution is a little stronger than O·lN.
The volume of water that must be added to make the solution
exactly 0·1000N is computed as de~cribed in Example 15 4
(Section I, 24). .
The standard solution thus prepared remains of the same
strength for a very long period if evaporation is prevented.
Ill, 32. Determination of silver in a silver alloy.-A silver
three-penny piece (or any other small silver coin) is suitable
for this determination. Clean the cOIn with emery cloth
and weigh it accurately. Place it in a ·250 ml. 'conical flask,
add 5 m!. of water and 10 ml. of concentrated nitric acid;
place a funnel in the mouth of the flask to avoid mechanical
loss. Warm the flask gently until the coin has dissolved.
Add a little water and boil Jor 5 minutes in order to expel
oxides of nitrogen. Transfer the cold solution quantitatively
to a 100 m!. standard flask and make up to the mark with
distilled water. Titrate 25 ml. portIons of the solution with
standard O·IN thiocyanate as described in Section m, 31.
Calculate the percentage of silyer in the alloy.
1 MI. N KCNS or NH 4CNS == 0·10788 g. -4g.
NOTE. The presence of metals whose salts are colourless does not
influence the accuracy of the determination, except that mercury
and palladium must be absent since their thiocyanates are insoluble.
Salts of metals (e.g., nickel and cobalt) which are coloured must not
be present to any considerable extent. Copper does-not interfere
provided it does not form more than about 40 per cent of the alloy.
m, 33. Determination of chlorides (Volhard's method).-
Discussion. The chloride solution is treated with an excess of
standard silver nitrate solution, and the residual silver nitrate
determined by titration ·with standard thiocyanate solution.
Now silver chloride is more soluble than silver thiocyanate and
would react with the thiocyanate thus:
AgCl + CNS- .= AgCNS Cl- +
(the theory of this process is given in Section I, 42B). It is
therefore necessary to remove the silver chloride by filtration.
The filtration may be avoided by the addition of a little nitro-
benzene (about 1 mL for each 0·05 g. of chloride); the silver
chloride is probably surrounded by a film of nitrobenzene.
For practice in this estimation, the student may determine
the HCI content of 'concentrated hydrochloric acid.
Procedure A (BCI content of concentrated hydrochloric
acid). Ordinary concentrated hydrochloric acid is usually 10
to lIN, and must be diluted first. Measure out accurately
10 ml. of the concentrated acid from a burette into a 1 litre
volumetric flask and make up to the mark with distilled water.
Shake well. Pipette 25 ml. into a 250 ml. conical flask, add 5
rul. of 6N nitric acid and then add, 30 ml. of standard O'IN
silver nitrate (or sufficient to give 2 to 5 ml. excess). Shake
to coagulate the precipitate, filter through a quantitative
filter paper (or through a Gooch or sintered glass crucible),
and wash thoroughly with very dilute nitric acid. Add 1 ml.
of the saturated ferric alum solution to the combined filtrate
and washings, and titrate the residual silver nitrate with
standard O'IN thiocyanate. Repeat the titration with two
other 25 ml. portions of the chloride solution.
Calculate the volume of standard O·IN silver nitrate that
has rested with the hydrochloric acid, and therefrom the per-
. centage of HCI in the sample employed.
Procedure B. Pipette 25 ml. of the diluted solution into a
250 ml. conical flask containing 5 ml. of 6N nitric acid. Add a
slight excess (2 to 5 ml.) of standard O'IN silver nitrate (about
30 ml.) from a burette. Then add 2 to 3 mL of pure (e.g.,
A.R.) Nitrobenzene and 1 ml. of the ferric alum indicator, and
shake vigorously to coagulate the precipitate. Titrate the
residual silver nitrate with standard O'IN thiocyanate until a
permanent faint reddish-brown colouration appears. Repeat
the titration with two other 25 ml. portions of the solution.
From the volume of silver nitrate solution added, subtract
the volume of silver nitrate solution that is equivalent to the
M
322 Quantitative Inorganic A~alysis
volume of standard thiocyanate required. Then calculate

the percentage of HCI in the sample.
1 Ml. N AgNO s = 0·03546 g. C1. = 0·03647 g. ~Cl.
m. 34. Determination of bromides (Volhard's method).-
Silver bromide is less soluble than silver thiocyanate so that it
is not necessary to filter off the silver bromide (compare
previous Section). The bromide solution is acidified with
dilute nitric acid, an excess of standard 0 ·IN silver nitrate
added, the mixture thoroughly shaken, and the residual silver
nitrate determined with standard O·lN thiocyanate, using
ferric alum as indicator.
1 MI. N AgNO a == 0·07992 g. Br s 0·08093 g. HBr.
HBr in the constant b.p. 'hydrobromic acid, or of bromine in
potassium bromide, preferably of A.R. quality (compare
Section UI, 23).
m, 35. Determination of iodides (Volhard's method).-
Here also there is no need to filter off the silver halide since
silver iodide is very much less soluble than silver thiocyanate.
In this estimation the iodide solution must be very dilute in
order to reduce adsorption effects. The dilute iodide solution
(ca. 300 mI.), acidified with dilute nitric acid: is treated ~ery
slowly and with vigorous stirring or shaking with standard
o'IN silver nitrate until the yellow precipitate coagulates
and the supernatant liquid appears -colourless. Silver nitrate
is then present in excess. I Ml. of saturated iron alum solution
is added, and the residual silver nitrate is titrated against
standard O·lN thiocyanate.
1 Ml. N AgNOs == 0.1269 g. I == 0·1279 g. HI.
HI in the constant boiling point hydriodic acid or of iodide in
potassium iodide, preferably of A.R. quality (compare Section
IIT,27). .
m, 36. Determination of hyposulphites. *-Discussion.
The method depends upon the reduction of ammoniacal silver
nitrate solution by the hypo sulphite to metallic silver; the
silver is dissolved in nitric acid and titrated against standard
thiocyanate solution, using ferric alum as indicator.
Na 2 S20 4 + 2AgNO a + 2HzO = 2Ag + 2NaHSO a + 2HNO a.
* Also called hydrosulphites.
The procedure has the advantage that the solid is weighed
directly into the reagent,. and thus avoids the rapid atmos-
pheric oxidation of the solution.
Procedure (hyposulphite content of commercial "hydros ").
Weigh out accurately about 0·4 g. of powdered" hydros "
(sodium hyposulphite, usually Na 2S2G"H 20) from a stoppered
weighing bottle into a dry beaker, and cover it with at least a
two-fold excess of ammoniacal silver. nitrate solution (pre-
pared by the addition of dilute ammonia solution in excess
of that required to re-dissolve the silver hydroxide first
formed). As the salt dissolves, it is at once oxidised and
metallic silver is precipitated. It is essential not to warm the
solution. When dissolution of the salt is complete, filter the
precipitated silver on asbestos in a Gooch crucible (or through
a sintered glass o.r porous porcelain crucible), and wash it with
dilute ammonia solution containing a little ammonium nitrate
to 'prevent peptisation of the finely-divided metal. Transfer
the crucible and contents to the beaker in which precipitation
took place, dissolve the metal in nitric acid, and boil the
solution to remove nitrous fumes. Filter the resultant solu-
tion through a sintered glass crucible, dilute the filtrate., add 1
ml. of saturated ferric alum solution, and titrate with stan-
dard O'IN thiocyanate.
Calculate the percentage of Na 2S2 0, in the sample.
1 Ml. N NH 4CNS (or KCNS) == 0·08706 g. Na 2S2 0 4 •
m, 37. Determination of cobalt.-Discussion. This

method is based upon the formation of the compound
[Co(C 6 H 6 N)J(CNS)2 (Sections I. 62L and IV. 33D). Excess
of a standard solution of ammonium thiocyanate is added to a
solution of the cobaltous salt containing a little pyridine,
the precipitated complex salt is filtered off, and the excess of
thiocyanate titrated with standard silver nitrate solution.
Procedure. The sample should cohtain 0·05 to 0'1 g. of
cobalt as cobaltous salt. Dissolve it in 150 ml. of water in a
250 mI. volumetric flask, acidify with dilute nitric acid (to
litmus), add 3 ml. of pure pyridine, and excess of standard
O'IN ammonium thiocyanate (I). The cobalt is precipitated
as [Co(C 6 H sN)J(CNS)2' which settles readily and filters
easily. Dilute to the mark, and mix thoroughly. Filter a
portion of the solution through a quantitative filter paper, and
discard the first few ml. of the filtrate. Transfer 50 m!. of the
filtrate to a beaker or conical flask" dilute to 100 mI., add 1 m!.
of concentrated nitric acid, and then a known excess of stan-
dard O;IN silver .nitrate. Add 5 .rol. of ferric indicator (2),
and titrate the excess of silver nitrate with O·lN ammonium
thiocyanate in the usual way (Section m, 31).
Calculate the weight of cobalt w from the formula:
w = [mt of NH,CNS X N - 5(mL of AgN0 3 X N)] 0·02947.
Notes. 1. Standardise the 0 ·IN ammonium thiocyanate against
standard O·IN silver nitrate: the latter may be standardised, if
necessary, against A.R. sodium chloride.
2. Dissolve 5 g. of A.R. iron alum in a mixture of 40 m!. of water
and 10 m!. of 6N nitric add.
m, 38. Determination of nicltel.-The procedure is similar
to that given in the preceding Section for Cobalt. The com- .
plex has the formula [Ni(CsHsN)J(CNS)t (compare Section
IV, 32C). The weight of nickel w is calC'ulated from the
formula:
w = [ml. of NH,CNS X N - 5(ml. of AgNOa X N)] 0·02935.
Bote. The method may also be applied to the determination
of copper (Section IV, 19E), cadmium (S~ction IV, 18D), and
zinc (Section' IV, 34D).
I
m, 39. Determination of fluoride as lead cWorofluoride.-
Discussion. This method is based upon the precipitation of
lead chlorofluoride PbCIF, which can be weighed as such, or,
more conveniently, the chlorine is determined by Volhard's
method from which the fluorine content can be calculated.
The advantages of the method are: the precipitate is granu-
lar, settles readily and is easily filtered; the fa~tor for con-
version to fluorine is low; the volumetric procedure is carried
out at pH 3·6 to 5·6, so that substances which might be
coprecipitated, such as phosphates, sulphates, chromates and
carbonates, do not interfere. Aluminium must De entirely
absent since even very small quantities cause low results; a
similar effect is produced by boron (> 0·05 g.). ammonium
(> 0·5 g.), and sodium or potassium (>10 g.). Iron must be
removed, but zinc is without effect. Silica does not vitiate
method, but causes difficulties in filtration.
fluorine in sodium fluoride. .
. Procfldure. Weigh out accurately about 1·5 g. of dry sodium
fluoride, dissolve it in water, and make up to 250 ml. in a
volumetric flask. Shake well. Pipette 25 m!. of the solution
Volumetric A nalys-is 325
into a 400 ml. beaker*, add 2 drops of bromo-phenol blue
indicator, 3 ml. of 10 per cent sodium chlor;de and dilute the
mixture to 250 m!. Add dilute nitric acid until the colour
just changes to yellow, and then add dilute sodium hydroxide
solution until the colour just changes to blue. Treat with 1
ml. of concentrated hydrochloric acid, then with 5·0 g. of
A.R. lead nitrate, and heat on the steam bath. Stir gently
until the lead nitrate has dissolved, and then immediately
add 5·0 g. of crystallised sodium acetate and stir vigorously.
Digest on the steam bath for 30 minutes, with occasional
stirring, and allow to stand overnight.
Meanwhile, a washing solution of lead chlorofluor~de is
prepared as follows. Add a: solution of 10 g. of lead nitrate
in 200 ml. of water to 100 ml. of a solution containing 1·0-g.
of sodium fluoride and 2 ml. of concentrated hydrochloric
acid, mix it thoroughly, and allow the precipitate to settle.
Decant the supernatant liquid, wash the precipitate by
decantation with 5 portions of water, each of abo~t 200 ml.
Finally add 1 'litre of water to the precipitate, shake the
mixture at intervals during an hour, allow the precipitate' to
settle, and filter the liquid. Further quantities of wash
liquid may be prepared as needed by treating the precipitate
with fresh portions of water. The solubility of lead chloro-
fluoride in water is 0·325 g. at 25°C. .
Separate the original precipitate by decantation through a
Whatman No. 542 or No. 42 paper. Transfer the precipitate
to the filter, wash once with cold water, four or five times with
the saturated solution of lead chlorofluoride and finally once
more with cold water. Transfer the precipitate and paper
to the beaker in which precipitation was made, stir the paper
to a pulp. in 100 ml. of 5 per cent nitric acid, and heat on the
steam bath until the precipitate has dissolved (5 minutes).
Add a slight excess of standard O'lN silver nitrate, digest on
the steam bath for a further 30 minutes, and allow to cool to
room temperature while protected from the light. Filter the
precipitate of silver chloride through a sintered glass crucible,
wash -with a little cold water, and titrate the residual silver
nitrate in the fj.ltrate and washings with standard O'lN
thiocyanate. Subtract the amount of silver found in the
filtrate f!,om that originally added. The difference represents
the amount of silver that was required to combine with the
chlorine in the lead chlorofluori9-e' precipitate.
'" The details apply to a volume of solution corresponding to 0·01 to 0·1 g.
of fi uorine.
.a26 Quantitative Inorganic. Analysis
Calculate the percentage of fluorine in the sample.
I Ml. N AgNO s := 0·01900 g. F.
m, 40. Determination of arsenic as silver arsenate.-
Discussion. Arsenates in solution are precipitated as silver
arsenate AgaAs04 by the addition of neutral silver nitrate
solution: the solution must be neutral or if slightly acid, an.
excess of sodium acetate must be present to reduce the acidity;
if strongly acid, most of the acid should be neutralised by
aqueous sodium hydroxide. The silver arsenate is dissolved
in dilute nitric acid, and the silver titrated with standard
thiocyanate solution. The silver arsenate has nearly 6
times the weight of the arsenic, hence quite small amounts
of arsenic may be determined by this procedure.
Arsenites must first be oxidised by treatment with nitric
acid. Small amounts of antimony and tin do not interfere,
but chromates, phosphates, molybdates, tungstates and
vanadates, which precipitate as the silver salts, should be
absent. An excessive amount of ammonium salts has a
solvent action on the silver arsenate.
A general method applicable to all arsenic compounds will
be described. ,
Procedure. Fuse 0·5 g. (or less, according to the amount of
As present) of the finely-powdered substance with 3 to 5 g.
of a I: 1 mixture of A. R. sodium carbonate and A.R.· potassium
nitrate, placing about one-third of the mixture on -top of the
charge, or, alternatively, use 10 to 12 times th~ weight of
sodium peroxide. Extract the cooled melt with hot water,
and filter. Acidify the filtrate, containing the soluble arsen-
ate, with dilute nitric acid, boil to expel carbon dioxide, and
cool. Add sufficient sodium hydroxide solution to give an
alkaline reaction to phenolphthalein, and then discharge the
red colour from the. solution by just acidifying with acetic
acid. Add a slight excess of silver nitrate solution with
vigorous stirring, and allow the precipitate to settle in the
dark. Pour off the supernatant liquid through a sintered
glass crucible, wash the precipitate by decantation with cold
distilled water, transfer the precipitate to the crucible, and
wash it free from silver nitrate solution. Wash out the
receiver thoroughly. Dissolve the silver arse]J.ate in dilute
nitric acid (ca. N) (which leaves any silver chloride undis-
solved), -w;ash with very dilute nitric acid, and make up the
filtrate and washings to 250 m!. in a volumetric flask. Titrate
a convenient aliquot portion with standard ammonium (or
Volume!rt'c Analysis 327
potassium} thiocyanate solution in'the presence of ferric alum
as indicator (Section m, 31). Calculate the percentage of
arsenic in the sample. .
1 Ml. N thiocyanate == 0·02947 g. of As.
m. 41. Determination of cyanides.-Discussion. The
theory of the titration of cyanides with silver nitrate solution
according to Liebig (1851) has been' given in Section I, 41.
All silver salts except the sulphide are readily soluble in excesS
of a solution of an alkali cyanide, hence chloride, bromide qnd
iodide do not interfere. The only difficulty in obtaining a
sharp end point lies in the fact that silver cyanide is often
precipitated in a curdy form which does not readily re-
dissolve, and, moreover, the end point is not easy to detect
with accuracy.
There are two methods for overcoming these disadvantages.
In the first the precipitation of silver argentocyanide at the
end point can be avoided by the addition of ammonia solution,
in which it is readily soluble:
Ag[Ag(CN)2] + 4NH 40H' 2[Ag(NH a)z]CN* + 4H 20.
1£ a little potassium iodide solution is .added before the titra-
tion is commenced, the sparingly soluble silver iodide, which
is insoluble in ammonia solution, will be precipitated at the
end point.
2[Ag(NH a}z]CN + 2KI + 4H zO
= 2AgI + 2KCN + 4NH 40H.
At the end point, the cyanide, in which silver iodide is soluble,
is no longer present to keep it in solution, and since, unlike
the silver argentocyanide, it is insoluble in ammonia solution,
silver iodide is precipitated. The precipitation is best seen
by viewing against a black background.
In the second method diphenylcarbazide is employed as an
adsorption indicator. The end point is marked by the pink
colour becoming pale violet (almost colourless) on the colloidal
precipitate in dilute -.solution (ca. O·OlN) before the opales-
cence is visible. In O·IN solutions, _the colour change is
observed on the precipitated particles of silver argentocyanide.
KCN in potassium cyanide (single salt).
liOTE. Potassium cyanide and all other cyanides are deadly poisons
and extreme care must be taken in their use.
• Or [Ag(NH3hJ+ + [Ag(CN).r.
Procedure A. (KCN in potassium cyanide). Weigh out
accurately abo~t 3·5 g. of potassium cyanide (single salt)
from a glass-stoppered weighing bottle, dissolve it in water
and make up to 250 mI. in a volumetric flask. Shake well.
Transfer 25 mI. of this solution by means of a burette and
NOT a pipette to a 250 ml. conical flask, add 75 ml. of water,
5 to 6 mI. of 6N ammonia solution, and 2 mI. of 10 per cent
potassium iodide solution. Place the flask on a sheet of
black paper, and titrate ~ith standard O·IN slIver nitrate.
Add the silver nitrate solution dropwise as soon as the yellow
silver iodide shows any signs of persisting. When 1 drop
pro.duces a permanen~ turbidity, the end point has been
reached. Repeat the titration with two other 25 ml. portions
of the cyanide solution.
Calculate the KCN content of the sample.
1 Ml. N AgN0 3 == 0·05204 g. CN "'" 0·1304 g. KCN.
Procedure B. Transfer 25 mI.. of the cyanide solution,
prepared as detailed in Procedure A, by means of a burette
and NOT a pipette into a 250 mI. conical flask, add two to three
drops of the diphenylcarbazide indicator, and titrate against.
the standard o·IN silver nitrate until a permanent violet
colour is just produced. Repeat the titration with two other
25 mI. portions of the cyanide solution. I
The diphenylcarbazide indicator is prepared by dissolving 0·1 g.

of diphenylcarbazide in 100 m!. of alcohol.
Calculate the percentage of KCN in the sample.
Ill, 42. Determination of nickel by potassium cyanide.-
Discussion. This ~ethod depends upon the fact that nickel
ions can react with pqtassium cyanide in slightly ammoniacal
solution to form a complex anion, [Ni(CN)J- - :
[Ni(NH 3)6JC1 2 + 4KCN + 6H 20
= K 2[Ni(CN)J + 6NH 40H + 2KCI.
If the ammoniacal solution of nickel contains a little silver
iodide in suspension and potassium cyanide solution is added,
the turbidity will not disappear until all the nickel has been
converted into the complex cyanide. The slightest excess of
cyanide is therefo.re indicated by the clearing of the liquid;
this -excess may be determined exactly by adding standard
silver nitrate solution until the turbidity is reproduced.
AgI + 2KCN K[Ag(CN)2J + KI ;
Ag+ + r = Agl.
Copper, cobalt and zinc, which form complex cyanides,
must be absent; the addition of alkali pyrophosphate, how-
ever, renders zinc' harmless. Aluminium, manganese and
iron are kept in solution by the addition of citric acid, tartaric
acid 'or sodium pyrophosphate:
The potassium cyanide solution is standardised against
standard O·lN silver nitrate as described in Section m, 41.
It is evident that:
. AgN0 3 == 2KCN == Nij2,
i.e., 1 ml. N AgNO s == 0·1304 g. KCN = 0·02935 g. Ni.
For practice in this determination, the student may'deter-
mine the percentage of nickel (a) in pure nickel ammonium
sulphate and (b) in a steel.
Procedure (Ni in nickel ammonium sulphate) Dissolve 14
g. of A.R. potas~ium cyanide in 1 litre of distilled water in·a
volum,etric 'flask, and shake well. Prepare a solution of 10
g. of A.R. potassium iqdide in 100 ml. of water. Standardise
the potassium cyanide solution against standard 'O'lN silver
nitrate (Section m, 41).
Weigh out accurately about 5 g. of pure nickel ammoniu~
sulphate NiSO,,(NH,)zSO,,6H 2 0·, dissolve it" in water con-
tai;ning a little dilute sulphuric acid, and make up to 250 ml.
in a volumetric flask. Shake well. Pipette 25 ml. of this
solution .into a 250 mI. conica:! flask, add 125 ml. of water,
followed by 6N ammonia solution until the fluid smells
slightly of it, and then add a further 3 ml. If a precipitate
forms, it will redissolve if a little ammonium chloride solution
is added. Then add 5 ,ml. of the potassium iodide solution
and a small measured volume,. say 0·5 to 1 ml., of the standard
-O'lN silver nitrate. The silver iodide produced should cause
the liquid· to become cloudy. Titrate with the standard
potassium cyanide solution from a burette until the turbidity
just disappears.. This will take place when the potassium
cyanide added is equivalent .to the nickel in the solution plus
the silver in the known volume (0·5 to 1 ml.) of silver nitrate
solution added. Add standard O'lN silver nitrate dropwise
from a burette until the cloudiness due to silver iodide reap-
pears. The excess of potassium cyanide solution, whic~ is
equivalent to the silver nitrate solution, is thus accurately
d~termined.
* The A.R. product may be obtained, e.g., from the Mallinckrodt Chemical
Works, U.S.A. See Appendix. Section A, 9.
M*
Repeat the titration with two other 25 ml. portions of the
nickel solution.
Calculate the volume of the standard potassium cyanide
solution which has reacted with the nickel, and therefrom the
percentage of nickel in· the sample.
Procedure (Ni in nickel steel). The most accurate method
for the analysis of any steel containing nickel is to precipitate
the nickel as the 9.imethylglyoxime complex, and to determine
the nickel in the latter.
Weigh out accurately about 1 g. of the nickel steel* into a
350 Il!l. conical flask, add 15 ml. of sN nitric acid, and place
a funnel in the mouth of the flask to prevent mechanical loss. '
When the first violent action is over, boil gently until all the
steel has dissolved, adding a little concentrated hydrochloric
acid if necessary. Continue the boiling for 5 minutes in
order to expel oxides of nitrogen. Dilute to 200 mI., and add
8 g. of pure tartaric acid. When all the tartaric acid has
dissolved, neutralise with concentrated ammonia solution, and
add ca. 1 ml. in excess. If any insoluble material is present,
filter it through a quantitative filter paper, and wash the
latter with hot water. Acidify the solution with hydrochloric
acid, heat to 80°C and add 20 to 25 ml. of a 1 per cent alcoholic
solution of dimethylglyoxime. Add ammonia solution until
the liquid is slightly alkaline, and allow to st~nd for 30 to 60
minutes, preferably on a water bath. Filter the precipit~te
through a Gooch or Jena sintered glass crucible, and wash the
precipitate thoroughly with hot water. Dissolve the precipi-
tate in 25 1111. of 6N nitric acid, add 20 ml. of 9N sulphuric
acid and boil for 20 minutes. Add 6 g. of citric acid, render
the solution slightly ammoniacal, and, when cold, titrate
the nickel as described above.
Calculate the percentage of nickel in the steel.
m, 43. Determination of thiocyanates.-Discussion. The
method (H. E. Williams, 1939) consists in the titration of the
thiocyanate in dilute nitric acid solution )Vith a standard
solution of mercuric nitrate, using a little iron alum solution
as indicator. Immediately the thiocyanate i~ completely
converted into the mercuric salt, the solution becomes water-
wh~te. The end point is very sharp.
The method may also be applied to the determination of
• Either Nirhdale's " Nickel steel, No. 22" (one of the Analysed Samples
for Students), or the Bureau of Analysed Samples" Nickel Steel, T " (a British
Chemical Standard) is suitable.
thiocyanate and chloride in admixture when the chloride is
present in comparatively small amount (say, 5 per cent of the
thiocyanate). The neutral solution containing the thiocya-
nate and chloride is treated with excess of bismuth nitrate
solution. The precipitated bismuth oxychloride is separated,
the filtrate acidified with dilute nitric acid, and the thiocya-
nate determined as above.
Procedure. The following solutions should be prepared:
N-Mercuric nitrate. 100·3 Grams of dry A.R. mercury are
heated in a 600 ml. beaker (fume cupboard) with 250 ml. of
distilled water, and concentrated nitric acid is added until all
the mercury is dissolved. The action of the acid is rather
energetic, especially when hot, and the acid should be added
a little at a time. The solution is boiled to drive off all
nitrous fumes, cooled, transferred to .a standardised litre
flask, and diluted to 1000 ml. The O'IN solution is prepared
by diluting the N solution. Since the mercury is pure and
can be accurately weighed, the solution requires no standar-
disation.
Mercuric sulphate may replace the nitrate, but offers no
advantage; mercuric chloride cannot be used.
Dilute nitric acid. 150 Ml. of concentrated nitric acid are
added to 400-500 ml. of distilled water, boiled to drive off
nitrous gases, and diluted to 1000 ml.
Iron alum solution. 25 Grams of A.R. iron alum are dis-
solved in water, and diluted to 500 ml.
The solution containing about 0·2 g. of thiocyanate is
diluted to 150-200 ml. with distilled water, and 10 ml. of the
dilute nitric acid and 2 ml. of the iron alum solution are added.
The solution is titrated against the O'IN mercuric nitrate
solution with. constant agitation until the red colour of the
ferric thiocyanate just disappears, leaving the solution water-
white.
1 Ml. O·IN Hg(NO a)2 == 0'005808 g. CSN.
Thiocyanate and chloride in admixture. A solution of bis-
muth nitrate is required and is prepared as follows. 10
Grams of pure crystallised bismuth nitrate and 50 ml. of
A.R. glacial acetic acid are warmed in a 200 ml. beaker, and
30 ml. of distilled water are added gradually until complete
dissolution has been effected, the solution being stirred be-
tween each addition. When the salt has completely dissolved,
the solution is diluted to 500 ml. with distilled water. 10 Ml.
of this solution are required to precipitate approximately
0·01 g. of chloride, calculated as sodium chloride.
The neutral solution containing approximately 0·2 g. of
thiocyanate ,and 0·01 g. of chloride (calculated as NaCI)
is diluted to 150 ml. with distilled water, and 10 ml. of the
bismuth nitrate solution are added. The mixture is shaken,
left for 10-15 minutes in the cold, filtered through a quanti-
tative filter paper, and the precipitate well washed. The
filtrate should give no precIpitate upon the addition of a few
m!. of the bismuth nitrate solution. The filtrate is then
acidified with dilute nitric _g,cid, and, after adding 2 m!. of
the iron- solution, titrated with O'lN mercuric nitrate solution
as above. The chloride may be determined by any of the
usual methods by dissolving the basic chloride on the filter
paper in dilute nitric acid.
m. 44. Determination of barium and of sulphates. Barium
solutions, weakly acid with hydrochloric acid, give a red
precipitate of barium" rhodizonate with a freshly-prepared'
II
solution of the sodium salt of rhodizonic acid (dihydroxy-

diquinoyl, C6 0 4 (OHh ) :
CO CO
OC
/ "'-C-ONa OC
/ "'-C-~ I
I II + BaCl2 - I II, Ba + 2NaCl.

OC C-ONa OC C-O/
"'-CO/ "'-/
CO
The solubility of the red compound is much greater than that
of barium sulphate, henc~ upon the addition of a sulphate
solution to a solution of a barium salt, the barium will be
completely precipitated when the red colour changes to colour-
less (in weakly acid solution) or to yellow (in neutral solution).
This forms the basis of a volumetric method for the deter-
mination of barium. To obtain the best results, it is advisable
to work with concentrated solutions (volume up to 10 m!.
- 0·2 to 0·5N).
The indicator is freshly prepared as required by dissolving 15 mg.
of sodium" rhodizonate " in 5 mI. of water. Ten drops are used
for each titration.
A. Barium. The concentrated solution of the barium salt
(5 to 10 m!.) is treated with 10 drops of indicator and 1 to 2
m!. of alcohol. It is then titrated with vigorous shaking
against (a) 0·2N potassium sulphate (prepared from the A.R.
Volumetric , Analysis 333
solid) until the colour changes sharply from red to yellow, or
(b) against 0·2N sulphuric acid until the colour changes from
red to colourless.
B. Sulphate. Sulphate in acid solution cannot be titrated
directly against a solution of a barium salt, because 'rhodi-
zonic acid is unstable and soon decomposes. A known excess
of barium chloride solution (0·2 to 0'5N) is added to the sul-
phate solution, and the excess of barium titrated as above.
Here also the solutions should be concentrated (0·2 to 0'5N)
and the volume small (10 to 25 ml.).
Excellent results may be obtained as follows. A known
volume of the neutral alkali sulphate solution {say, 10 to 20
ml.) is placed in a glass-stoppered conical flask, 1 g. of pure
ammonium chloride added, followed by a known excess of
0·2N barium chloride (prepared from the A.R:- salt) -and 10
drops of the freshly-prepared indicator. The mixture is
thoroughly shaken, and 3 drops of 10 per cent hydrochloric
acid added; the further addition of one-third the volume of
alcohol improves the end point. The mixture is then titrated
against 0·2N potassium sulphate until the colour begins to
fade. The alkali SUlphate solution is then addert-dropwise,
with vigorous shaking and waiting for a short time between
the drops, until the red coI-our has completely disappeared.
The titration is repeated until consistent results are obtained.
By subtracting the volume of 0·2N potassium sulphate
from that of the 0·2N barium chloride originally employed,
the volume of the latter equivalent to the sulphate present in
the initial solution is found, since Ba =:= S04'
OXID~TION-REDUCTION TITRATIONS. OXIDIMETRY
AND REDUCTIMETRY
m, 45. General discussion.-The student should be familiar
with the general theory of oxidation-reduction given in
Chapter I; in particular, he should read:
Section I, 23-Equivalent weights. Normal solutions.
Sections I, 43 to I, 53-Theory of oxidation-reduction
reactions (including potentiometric titration).
With this groundwork, he should experience no difficulty
'in carrying out oxidation-reduction titrations intelligently
and, when once he has had practice in the fundamental simple
titrations, the student should be in a position to apply his
combined theoretical and practical knowledge to more
complex cases and to various practical problems.
The standard solutions of oxidising agents employed in the
titration of reducing agents include those of potassium per-
manganate, potassium dichromate, ceric sulphate, manganic
sulphate, iodine, potassium iodate, potassium bromate, and
chloramine-To The standard solutions of reducing agents,
used in the titration of oxidising agents, include those of
sodium thiosulphate, arsenious oxide, sodium oxqlate, oxalic
acid, ferrous salts, titanous sulphate or chloride, chromo us
sulphate or chloride. and vanadous sulphate or'chloride. The
mode of use and applications of each of these will be given in
the following Sections.
OXIDATIONS WITH POTASSIUM PERMANGANATE

ill, 46. Discussion.- This valuable and' powerful oxidising
agent was first introduced into volumetric analysis by F.
Margueritte in 1846 for the titration of ferrous iron. In
acid sol1ftions, the reduction can be represented by the follow-
ing equation:
MnO ..- + 8H+ + 5 £ ~ Mn++ \ + 4H 2 0,
from which it follows that the equivalent weight is one fifth
of the formula weight (1/5 mol)*, or 158'03}5, or 31·606.
• For elementary students. the equivalent weight may be deduced from the
hypothetical equation : ~
2KMnO. + aR,so. = K,SO. + 2MnSO. + aR.O + 50.

It follows that the equivalent weight is one fiftl! of the molecular weight
(KMnO./5).
334
Sulphuric acid is the most sUitable acid as it has no action
upon permanganate in dilute solutitm. With hydrochloric
acid, there is the likelihood of the reaction:
'2KMnO, + 16HCI = 2KCI + 2MnC1 2 + 5C12 + 8H 20,
taking place, and some permanganate may be consumed in
the formation of chlorine. This' reaction is particulariy
liable to occur with iron salts ,unless special precautions are
a
adopted (see below). With small excess of free acid, a very
dilute solution, low temperature and slow titration with
constant shaking, the danger from this cause is minimised.
There are, however, some titrations, such as those with
arsenious oxide, trivalent antimony and hydrQgen peroxide,
which can be carried out with hydrochloric acid.
In the analysis of iron ores, solution is frequently effected
in concentrated hydrochloric acid, the ferric iron is reduced
with a slight excess of stannous chloride solution, and the
excess of the latter removed w.ith mercuric chloride solution
(Zimmerman-Reinhardt process): the ferrou~ iron is then
determined. in the resultant solution. To do this, it is best
to add about 25 ml. of Zimmermann and Reinhardt's solution,
which is prepared by dissolving 50 g. of crystallised manganese
sulphate MnS0 4 ,4H 2 0 in 250 ml. of water, adding a cooled
mixture of 100 mI. of c~ncentrated sulphuric acid and 300 ml~
of water, followed by 100 ml. of syrupy phosphoric acid. The
manganous sulphate "lowers the oxidation potential of the
permanganate (compare Sections 1,46,47 and Fig. 1-17) and
thereby makes it a weaker oxidising agent; the tendency of
the permanganate to oxidise chloride ions is thus reduced.
The manganous ion also acts as a catalyst in promoting the
speed of the reduction of the permanganate ion, and as a
negative catalyst for the oxidation of the chloride ion. The
phosphoric acid combines with tbe yellow ferric ions to form
colourless complex ions, thus rendering the end point' more
clearly visible. The main purpose of the phosphoric acid is,
however, to lower the oxidation potential of the ferric-ferrous
system (Section I, 46) and 'thus to increase the reducing power
of the ferrous iron.
For the titration of colourless or slightl? coloured solutions,
the use of an indicator is unnecessary, since as little as 0·01 ml.
of O·OIN potassium permangariate imparts a pale pink colour
to 100 ml. of water. The intensity of the colour in dilute
solutions may be enhanced, if desired, by the addition of
an Gxidation-reduction indicator, such as diphenylamine,
diphenylbenzidine, sodium diphenylamine sulphonate, tri-
ortho-phenanthroline ferrous sulphate or N-phenylanthranilic
acid, near the end point of the reaction; this is usually
not required. .
Potassium permanganate also finds considerable application
in strongly alkaline solutions. -Here two consecutive partial
reactions take place: (i) 'the relatively rapid reaction
MnO",- + E = Mu.O",- -,
and (ii) the relatively slow reaction
Mn0 4- - + 2H zO + 2E -; Mn02 + 40H-.
By suitably controlling the experimental conditions, it is
possible to arrange that reaction (i) occurs almost exclusively;
the equivalent weight is then 1 mol. Excess of deci-molar
potassium permanganate which is 1 to 2N with respect to
sodium hydroxide i.s added to a. solution of the substance to
be oxidised. After 10 to 20 minutes the excess of perman-
ganate is back-titrated either in an alkaline medium with a
standard solution, of sodium formate
(_2~nO,- + 30H- + HC0 2- = 2MnO,- - + C0 3-: - + 2H 2Q)
or in an acid medium with oxalic acid. *
Potassium permanganate is not a primary standard. It is
• difficult to obtain the substance perfectly pure and free from
manganese dioxide. Moreover, ordinary distillJd water is
likely to contain reducing substances (traces of organic mat-
ter, etc.) which will react with the potassium permanganate to
form manganese dioxide. The presence of the latter is very
objectionable because it catalyses the auto-decomposition
of the p.ermanganate solution on standing. For these reasons,
potassium 'permanganate solution is never lllade up exactly
by weight, but after removal of the manganese dioxide by
filtration through purified asbestos or glass wool o~ through a
sintered glass filter crucible (Jena, No.4) is standardised by
one of several methods. These methods include the use of
the following substances: sodium oxalate Na 2C20" arsenious
oxide As 20 3 , metallic iron, fer:rous ammonium sulphate
FeSO,,(NH,)zSO,,6H zO, oxalic acid H ZC20,,2H zO, and pbtas-
. sium tetroxalate KHCz04,HzC204, 2H2 0.
Of these substances sodium oxalate was formerly regarded
as the most trustworthy, since it is readily obtained pure and
* For further experimental details and an account of the numerous applica-
tions of this method. the reader is referred to W. Bottger-R. E. Oesper.
Newer Methods of Volumetric Chemical Analysis. 1938. p. 55. (D. Van
Nostrand; Chapman and Hall.)
anhydrous. The U.S. Bureau of Standards supplies a pro-
duct of 99·95 per cent guaranteed purity, and the ordinary
A.R. substance has a purity of at least 99·9 per cent.* The
experimental procedure hitherto employed was due to R. S.
McBride (1912). A solution of the oxalate, acidified with
dilut~ sulphur~c acid and warmed to 80° to 90°C, was treated
slowly (10 ·to 15 ml. per minute) and with constant stirring
with the permanganate solutio!1 'Until the first permanent
faint pink colour was obtained; the temperature near the end
point was not allowed to fall below 60°C. R. M. Fowler and
H. A. Bright (1935) have, however, shown that with McBride's
procedure the results may be 0·1 to 0'i!5 per cent high; the
titre depends upon the acidity, the temperature, the rate of
addition of the permanganate solution, and upon the speed of
stirring. These authors recommend a more rapid addition
of 90 to 95 per cent of the permanganate solution (about 25
to 35 m!. per minute) tq a solution of sodium oxalate in 2N
sulphuric acili at 25° to 30°C, the solution is then warmed to
55° to 60°C and the titration completed, the last 0·5 to 1 m!.
portion being added dropwise. The method is accurate to
0·06 per cent. Full experimental details are given in Proce-
dure B below.
2KMnO, + 3H 2SO, + 5H2 C20,
;= MnSO,+K 2 S04 + 10C0 2 + 8H 2 0.
The procedure of H. A. Bright (1937), which utilises arseni-

ous oxide as a primary standard and potassium iodide or
potassium iodate as a catalyst for the reaction, is more con-
venient in practice and promises to take first place as a trust-
worthy method for the standardisation of permanganate
solutions: Like sodium oxalate, A.R. arsenious oxide has
a purity of at least 99·9 per cent, and the U.S. Bureau of
Standards supplies a product of 99·98 per cent guaranteed
purity.t The results by this method agree to within 1 part
in 3000 with the sodium oxalate procedure of Fowler and
Bright. Full experimental details are given in Procedure A.
5As 20 3 + 4MnO,- + 12H+' = 5As 20 S + 4Mn++ + 6H 2 0;
5As20 3 + 4KMnO, + 12 HCI
= 4MnCl 2 + 4KCI + 5As 2 0 S + 6H 2 0
... A.R. sodium oxalate of the Mallinckrodt Chemical Works, U.S.A., has
an assay value of 99 ·95 per cent. The Bureau of Analysed Samples of Middles-
brough, England, supply a .. British Chemical Standard" sodium oxalate of
99·94 per cent purity. See also Appendix, Section A, 9.
t A.R. arsenious oxide of the Mallinckrodt Chemical Works, U.S.A., has
an assay value of 99'95 per cent. See Appendix, Sections A. 8, II;
Iron wire of 99·9 per cent purity is,available commercially.
The A.R. reagent* is a suitable standard, particularly if the
potassium permanganate solution is subsequently to be
employed in the estimation of iron. If the wire exhibits' any
sign of rust, it should be drawn between two pieces of fine
emery cloth, and then wiped with a clean dry cloth before
use. The reaction which occurs is. :
2KMn04 + 10FeS0 4 + 8H 2SO,
= K 2SO, + 2MnSO, + 5Fe 2(S04)3 + 8Hll:-O.
The use of crystallised oxalic acid, potassium tetroxalate,
and fe.rrous ammonium sulphate (Mohr's salt) for 5tandardisa-,
tion is not recommended for accurate work; they may,
however, be employed for elementary students in work in
which great accuracy is not required. All these substances
are hydrated, and unless special precautions are taken in
drying and storage the water content may be open to sus-
picion. Furthermore, for potassium tetroxalate and Mohr's
salt it is difficult to be quite sure that the constituent· salts are
present in exactly equimolecular proportions. The A.R.
substances have the following assay valuest: H 2C2 0,,2H 20 <t:
99·8 per cent; KHC 2 0"H 2C20 4, 2H 20 <t: 99·9 per cent; and
FeS04,(NH4)2S0" 6H 2 0 99'5 to 100'5' per cent. Apart from
the objections already mentioned, it is evidentl that A.R.
ferrous ammonium sulphate cannot be used directly for
standardisation, even for work of moderate accuracy. It
must first be dissolved in warm water, and the solution rapidly
cooled to ensure the production of small crystals; these are
filtered at the pump, washed with a little distilled water,
repeatedly pressed between sheets of clean filter paper, and
.placed in a current of air to dry thoroughly. The 'Substance
is dry when the small grains no longer adhere to each other or
to the paper in which they are contained, and is then preserved
in a stoppered bottle for use.
m, 47. Preparation of O·IN potassium permanganate.-
Weigh out about 3·2 to 3·25 g. of A.R. . potassium perman-
• The author has found that the" Iron Wire Analytical Reagent" su~plied
by the Mallinckrodt Chemical Works, U.S.A.,in sealed glass vials to be very
satisfactory. The assay value is 99'9 per cent Fe and the percentages of
impurities are given on the label. If the vials are carefully resealed with wax
after each opening, and care be taken to prevent the introduction of moisture
from the hands and other sources, the iron wire retains its original bright
surface for long periods.
• These figures are taken from A nalaR Standards jor Laboratory Chemicals,
2nd edition, 1937, published jointly by Hopki~ and Williams Ltd., London,
and ~ritish Drug Houses Ltd., London. ,
V olumetr£C Analysis 339
ganate on a watch glass, transfer it to a 1500 ml. beaker, add I
litre of water, cover the beaker with a clock glass (or Fisher
" speedyvap" beaker cover), heat the solution to boiling,
boil gently for 15 to 30 minutes, and allow the solution to
cool to the laboratory temperature. Filter the solution
through a funn~l containing a plug of purified glass wool, or
through a Gooch crucible provided with .a pad of purified
asbestos, or, most simply, through a Jena sintered glass filter
crucible or funnel. Collect the filtrate in a vessel which has
been cleaned with chromic acid mixture and then thoroughly
washed with distilled water. The filtered solution should be
stored in a clean glass-stoppered bottle, and kept in the dark
or in diffuse light except when in use: alternatively, it may be
kept in a bottle of dark-brown coloured glass.
m, 48. Standar~ation of permanganate solutioDS.-

Procidure A. With arsenious· oxide (H. A. Bright, 1937).
Dry some A.R. arsenious oxide at 105° to 110°C for 1 to 2
hours, and allow to cool in a desiccator. Accurately weigh
approximately 0·25 g. of the dry oxide, and transfer it to a
400 mI. beaker. Add 10 ml. of a cool solution ·of sodium
hydroxide, prepared from 20 g. of sodium hydroxide and 100
m!. of water (I). Allow to stand for 8 to 10 min-q.tes, stirring
occasionally. When solution is complete, add 100 ml. of
water, 10 m!. of pure concentrated hydrochloric acid, and 1
drop of 0-0025M potassium iodide or potassium iodate (2).
~dd the permanganate solution from a burette until a faint
pink colour persists for 30 seconds_ Add tne last 1 to 1-5 ml.
dropwise, allowing each drop to become decolourised before
the next drop is introduced. (For the most accurate work it
is necessary to determine the volulJle of permanganate solu-
tion required ,to duplicate the pink colour at the end point_
This is done by adding permanganate solution to a, solution
containing the same amounts of alKali, aGid cmd catalyst as
were used in the test_ The correction should not be more
than 0·03 ml.*) Repeat the determination with two other
similar quantities of-arsenious oxide_ Calculate the normality
of the potassium permanganate solution_ Duplicate deter-
minations should agree Within 0-1 per cent .
... The end point can also be determined with tri-ortho-phenanthroline
ferrous sulphate as indicator_ As the end point iii approached one drop of
the indicator (Section m, 63) is added:" The permanganate solution is then-
added dropwise until the pink colour of the indicator changes to a very faint
blue. The blank correction should average about 0-02 rol.
340 Quafttitative Inorganic Analysis
Notes. 1. For elementary stt/dents, it is sufficient to weigh out
accurately about 1·25 g. of A.R. arsenious oxide, dissolve this in
50 m!. of a cool 20 per cent solution of sodium hydroxide, and make
up to 250 m!. in a volumetric flask. Shake well. Remove 25 m!.
of this solution by means of a burette and-not with a pipette (caution
-the solution is highly poisonous), transfer to a 250. to 350 m!. conical
flask, add 100 m!. of water, 10 m!.l)f pure concentrated hydrochloric
acid, 1 'drop' of potassium iodide solution, and titrated with the
.permanganate solution to the first permanent pink colour as detailed
above. Repeat with two other 25 ml. portions of th~ solution.
Successive titrations should agree within 0·1 ml.
2. 0'0025M potassium iodide = 0·41 g. KI per litre; 0'0025M
potassium iodate = 0·54 g. KIO a per litre.
Calculation. It is evident from the equation given in
Section m, 46 and also from the equation
AS 20 3 + 20 = AS 20 5
that the equivalent weight of arsenious oxide is one quarter

of the molecular weight, 197 '82/4 or 49·46 g. 1 Ml. of a normal
solution contains the milli-equivalent weight, or 0·04946 g.
If the weight of arsenious oxide be divided by the number of
ml. of potassium permanganate solution to which it is equiva-
lent as found by titration, we have the -weight of primary
standard equivalent to 1 ml. of the permanganate solution.
If this l~st value be divided by the milli-equivalent of arseni-
ous oxide, the normality of the permanganate solution is
obtained (for details of an actual calculation, see under
Sodium Carbonate, Section m, 4A).
In the method' described in Note 1, the normality of the
arsenious oxide solution is calculated from the weight of
C),rsenious oxide used (1 litre of a N solution contains 49·46
g. of As 2 0 a) •• The normality of the permanganate solution is
'then computed from the relation ~
VA X nA = VB X nB
(compare Section m, 4A).
Procedure B. With sodium oxalate (R. M. Fowler and H.
A. Bright, 1935). Dry some A.R. sodium oxalate at 105 0
to 1l0°C for 2 hours, an.d a1low it to cool in a desiccator.
Weigh out accurately from a weighing bottle about 0·3 g.
of the dry sodium oxalate into a 600 m!. beaker, add 240 m!. of
water, and 12·5 m!. of concentrated sulphuric acid (caution)*.
* Or 250 ml. of approximately 2N sulphuric acid, prepared by adding 50 ml.
of concentrated sulphuric acid cautiously to 950 mI. of water.,
Volumetric Analysis au
Cool to 25° to 30°C, and stir uI}.til the oxalate has dissolved
(1). Add 90 to 95 per cent of the .required quantity of per-
manganate solution from a burette at a rate of 25 to 35 ml.
per minute while stirring slowly (2). Heat to 55° to 60°C
(use a thermometer as stirring' rod), and complete the titration
by adding permanganate solution until- a faint pink colour
persists for 30 seconds. Add the last 0·5 to 1 ml. dropwise
with particular care to allow each drop to become decolourised
before the next is introduced. (For the most exact work, it is
necessary to determine the excess of permanganate solution
required to impart a pink colour to the solution. This is
done by matching the colour produced by adding permangan-
ate solution to the same volume of boiled and cooled diluted
sulphuric acid at 55° ~o 60°C. This correction usually
amounts to 0·0'3 to 0·05 ml.) Repeat the determination
with two other similar quantities of sodium oxalate.
Calculate the normality of the potassium permanganate
solution. Duplicate determinations should agree within 0·1
to 0·2 per cent.
Notes. 1. For elementary students, it is sufficient 'to weigh out
accurately about 1·7 g. of A.R. sodium oxalate, transfer it to a 250
m!. volumetric flask, and make up to the mark. Shake well. Use
25 ml. of this solution per titration and add 150 m!. of ca. 2N
. sulphuric acid. Carry out the titration rapidly at the ordinary
temperature until the first pink colour appears throughout the
solution, and allow to stand until the solution is colourless. Warm
the solution to 50° to 60°C and continue the titration to a permanent
faint pink colour. It must be remembered that oxalate solutions
attack glass, so that the solution should not be stored more than a
few days.
2. An approximate value of the volume of permanganate solution
required can be computed from the weight of sodium oxalate em-
ployed. In the first titration about 75 per cent of this volume is
added, and the determination is completed at 55° to 60°C. There-
after, about 90 to 95 per cent of the volume of permanganate
solution is added at the laboratory temperature.
Calculation. This is similar to that described under
Procedure A. The equivalent weight of sodium oxalate is
1/2 mol, or 67 ·01 g.
1 Ml. N Na2 C20" == 0·06701 g. Na 2C20 4 •
Procedure C. With metallic iron. Use A.R. iron wire of
99·9 per cent assay value (see last but one footnote in Section
m, 46). Fit a tight rubber stopper with a bent delivery tube
into a 350 ml. conical flask and clamp it in a retort stand in an
inclined position, the tube being bent so as to dip into a small
beaker containing saturated sodium bicarbonate solution or
20 per cent potassium bicarbonate
solution (prepared from the A.R.
solids) (Fig. 3-4). Place 100 m!.
of 3N sulphuric acid (from 92 ml.
of water and 8 ml. of concentrated
sulphuric acid)' in the flask, and
add 0·5 to 1 g. of A.R. sodium
bicarbonate in two portions; the
carbon dioxide produced will drive
out the air. Meanwhile, weigh out
accurately about 0·15 g. of the
iron wi~e, place it quickly into the
flask, replace the stopper and bent
tube, and warm gently until the
FI~. 3-4. iron has dissolved completely. Cool
the flask rapidly under a stream
of cold water,· and then run in the permanganate solution
cautiously from a burette, with constant shaking, until the
faint pink colour is permanent. The addition of about 5 m!.
of pure syrupy phosphoric acid facilitates the detection of
the 'end point. Repeat the determination with two other
samples of the iron wire.
The reaction is :
2KMn0 4 + 10FeS04 + 8H2S04
= K 2S0 4 + 2MnS0 4 + 5Fe2 (S04)3 + 8H 20.
Calculation. The weight of iron wire taken, multiplied by
'()'999, will give the weight of pure iron used in each titration.
1 Ml. N KMn0 4 == 0·05584 g. Fe.
Procedure D. With ferrous ammonium sulphate. Weigh out
accurately about 9·8 g. of the A.R,salt: this should preferably
be recrystallised as detailed in Section m, 46. Transfer to a
250 ml. volumetric flask, ,add 200 ml. of water and 5 ml. of
concentrated sulphuric acid (or 200 ml. of N sulphuric acid).
Shake until the solid has dissolved and make up to the mark
with distilled water. Shake well. Remove 25 ml. of the
solution with a pipette, add 25 ml. of N sulphuric acid, an9-
titrate with the permanganate solution to the first permanent
pink colour. Here also the addition of a few ml. of syrupy
• As the flask cools. the solution is automatically drawn in until the pressure
of the carbon dioxide inside the flask is equal to the atmospheric pressure.
phosphoric acid improves the colour change at the end point.
Carry out 2 or at most 3 tit rations. These should not differ
?y more than 0·1 ml.
Calculation. The equivalent weight of. ferrous ammonium
slllphate in this reaction is, the molecular weight or 392·13
(2FeO +'0 = Fe z0 3). Compute the normality of the potas-
sium permanganate solution from the relation:
(Section 1;24).
Procedure E. With oxalic acid or potassium tetroxalate.
Weigh out accurately either about 1·6 g. of A.R. H 2CZO"
2H 20 or about 1·6 g. of A.R. KHCz04,H2C204,2H20, and
make up to 250 ml. in a volumetric flask. Shake well.
Remove 25 ml. portions with a pipette, add 150 mt of ca. 2N
sulphuric acid, and titrate against the permanganate solution
as detailed under Procedure B (sod}um oxalate). Successive
tit rations should agree within 0'1 ml.
CalQulation. The partial equation for both reactions is :.
C20 4- - + 2e = 2C0 2,
hence the equivalent weight of H 2CZO(,2H 20 is '0·5 mol or
63'03, and that for' KHCz04,HzC204,2H20 is 0·25 mol or
63·55. Calculate the normality of the potassium perman-
ganate solution as already detailed.
m~ 49. Perinanence of potassium permanganate solutions.

...LA pure permanganate solution, provided it is kept free from
dust, organic matter, precipitated manganese dioxide a!,1d
reducing vapours, will keep for a long time. For very accu-
rate work, however, it is advisable to standardise the solution
frequently, preferably with arsenious oxide. If filtered
solutions of permanganate are to be diluted, it is best to use
water which has been redistilled from alkaline permanganate,
e.g., conductivity water (Section II, lOC). The addition of
10 g. of potassium hydroxide per litre increases the stability
of the solution.
The student is reminded that owing to the decomposition
of perman'ganate solution by organic matter of any kind, it
cannot be filtered through filter paper, nor can it be used in a
clip burette because it is decomposed by the india-rubber
tube. Filtration is carried out through purified glass wool,
sbestos or a sintered glass funnel.
ANALYSES INVOLVING THE USE OF STANDARD

POTASSIUM PERMANGANATE SOLUTIONS
m, 50. Determination of ferrous iron.-The detailed
experimental method has already been given under ~rocedures
C and D of Section m, 48; the solution 'is_ acidified with dilute
sulphuric acid. If chloride ion is present high results are
obtained, because the reaction between ferrous iron and
permanganate induces the oxidation of hydrochloIic acid to
chlorine (or to hypochlorous acid). The chloride ion is
rendered almost harmless by the addition of a manganous
salt, preferably in the form of the so-called Zimmermann-
Reinhardt solution (Section m, 46). and by slow titration.
As an exercise, the student should determine the percentage
of iron in crystallised ferrous sulphate, FeSO,,7H 2 0. Weigh
out accurately about 7 g. of ferrous sulphate crystals, dissolve
in 200 ml. of N sulphuric acid (195 ml. of water and 5 ml. of
concentrated sulphuric acid) in a 250 ml. volumetric flask,
and make up to the mark with distilled water. Shake. well.
Remove 25 ml. of the solution with a pipette into a 250 mI.
conical flask, add 25 ml. of N sulphuric acid, and titrate with
standa~d O·IN potassium permanganate to the first perma-
nent faint pink colour. Repeat with two other 25 ml. portions
of the ferrous solution. Successive titrations should agree
within 0·1 ml.
Calculate the percentage of ferrous iron in the crystals.
2KMn04 + 10FeS04 '+ 8H 2S0 4
= K 2SO, + 2MnSO, + 5Fe2(S04)3 + 8H 0,
2
1 Ml. N KMnO, == 0·05884 g. Fe.
rn, 51. Reduction of ferric to ferrous iron.-Iron often

occurs in the ferric state, or as a mixture of ferrous and ferric
iron in ores and other materials. For the volumetric'deter-
mination of iron, the ferric iron must first be quantitatively
reduced to the ferrous state. The following methods, in de-
creasing order of importance and application, are available;
(A) with stannous chloride; (B) with amalgamated zinc
or cadmium, preferably in the Jones reductor; (0) with
titanous salts; (D) with sulphurous acid; (E) with hydrogen
sulphide; and (F) with zinc or magnesium and dilute sul-
phuric acid. Methods A and B are widely used. Method C
is described in Section m, 86. Method F has been largely
displaced by B:
A. Reduction with stannous chloride solution. Many iron
ores are brought into solution with concentrated hydrochloric
acid. In such cases the reducing agent frequently employed
is stannous chloride:
2FeCl a + SnC1 2 = 2FeC12 + SnCi 4 •
The hot iron solution (70 0 to 90 0 q, from about 0·3 g. of
iron ore, which should occupy a volume of 25 to 30, m!. * and
be 5 -to 6N with respect to hydrochloric acid, is reduced by
adding concentrated stannous chloride solution dropwise
from a separating funnel or a burette, with stirring, until the
yellow colour of the solution has nearly disappeared., The
reduction is then completed by diluting the concentrated
solution of stannous chloride with 2 yolumes of dilute hydro-
chloric acid, and adding the dilute solution dropwise, with
agitation after each addition, until the liquid has a faint green
colour, quite free from any tinge of yellow. The solution is
then rapidly cooled under the tap, with protection from the
air, and the slight excess of stannous chloride present removed
by adding 10 m!. of a saturated solution (ca. 5 per cent) of
mercuric chloride rapidly in one Hortion'. A slight silky white
precipitate of mercurous chloride should be obtained:
2HgCl 2 + SnCl 2 = Hg 2 Cl 2 + SnCI4 •
The oxidising agent has no appreciab)e effect upon the small
amount of mercurous chloride in suspension. If a heavy
precipitate forms or a grey or black precipitate is obtained,
too much stannous chloride has been used; the results are
inaccurate and the reduction must be repeated. Finely
divided mercury is oxidised by potassium permanganate (and
by potassium dichromate).
After the addition of the mercuric chloride solution, the
.whole is allowed to stand 5 minutes, and transferred quantita-
tively to a 600 m!. beaker or casserole with the aid of about
JOO m!. of water. 25 Ml. of the Zimmermann-Reinhardt
solution (the so-called" preventive solution ") are added, and
the mixture titrated slowly and with constant stirring with
standard O·IN potassium permanganate. The end point is
not the permanent pink colour which is obtained in the
absence of a chloride, but is the first pink colouration seen
throughout the whole solution; this colouration fades after
about 15 seconds. The fading is due to the reaction between
the permanganate "Solution and the mercurous chloride, and
• If the volume is greater than tllis, it must be reduced by careful evaporation
346 Quantitatt've Inorganic Analysis
also to the following reaction, which is, however, very slow in
cold acid solution:
2KMn0 4 + 3MnS0 4 + 2H 2 0 = 5Mn0 2 + K 2 SO, + 2H 2 S0 4 •
Blank runs on the reagents should be carried through all
the operations, and corrections made, if necessary.
The concentrated solution of stanhous chloride -is prepared by
dissolving 12 g.' of pure tin or 30 g. 'of A.R. crystallised stannous
chloride (SnC12 • 2H2 0) in 100 ml. of concentrated hydrochloric acid
and di~uting to 200 ml. with water.
Details of the Zimmermann-Reinhardt solution are given in
Section m, 46.
B. Reduction with amalgamated zinc or cadmium. The
Jones reductor. Both a:malgamated zinc a..nd cadmium are
excellent reducing agents for ferric iron:
2 Fe+++ + Zn = 2Fe++ + Zn++.
The former is generally employed. Zinc can be purehased
in the pure state: A.R. granulated zinc contains less than
0·002 per cent of iron. Both pure cadmium and zinc react
rather slowly with acids, b!lt upon treatment with a dilute
solution of a mercuric salt, the metal is covered with a thin
layer of mercury; the amalgamated metal reacts quite readily.
Reduction with amalg5Lmated zinc
is usually. carried out in the t
"reductor". due to C. Jones (1888- 6 cm.
1889). This consists of a column
of amalgamated zinc contained in a
long glass tube provided with a
stop-cock, through which thesolution
to be reduced may be drawn. A large 36cm_
1 I"ternal
dIameter
16·f9mm.
surface is exposed, and consequently
such a zinc column is much more
efficient than pieces of zinc placed in
the solution.
A suitable form of the Jones TrJ pump
reductor, with approximate dimen-
sions, is shown in Fig. 3-5. A
perforated porcelain plate, covered
with purified asbestos or glass wool, 22 em.
r-
I'- - "'- '
supports the zinc column. The tube
below the tap passes through a
tightly-fitting one-holed rub b e r
stopper into a 750 m!. filter flask. FIG. 3-5.
It is advisable to connect another filter flask in series
with the water pump, so that if any water" sucks back" it
will not spoil the determination. The amalgamated zinc is
prepared as follows. About 300 g. of A.R. granulated zinc
(or zinc shavings, or pure 20 to 30 mesh zinc) are covered with
2 per cent mercuric chloride solution in a beaker. The mix-
ture is stirred for 5 to 10 minutes, then the solution is decanted
from the zinc, and washed three times with water by decan-
tation. The resultant amalgamated zinc should have a
bright silvery lustre. The porcelain plate is placed in posi-
tion, covered with a layer of purified asbestos or glCl-sS wool,
and then the amalgamated zinc added; the latter should
reach to the shoulder of the tube. The zinc is washed with
disti.lled water (500 ml.), using gentle suction. If the reductor
is not to be used immediately, it must be left full of water in
order to prevent the formation of basic salts by atmospheric
oxidation, which impair the rt'!ducing surface.
To use the red.uctor for the teduction of ferric iron, proceed
as follows. The zinc is activated by filling the cup (which
holds abOut 50 ml.) with 2N (ca. 5 per cent) sulph.uric acid, the
tap being closed. The flask is connected to a filter pump, the
tap opened, and the acid slowly drawn through the column
until it has fallen}ust above the level o/the zinc; the tap is then
'closed and the process repeated twice. The tap is shut, the
flask detached, cleaned and replaced. The reductor is now
ready for use. It is important to note that during use the
level of the liquid should always be just above the top of the
zinc column. The solution to be reduced should have' a
volume of 100 to 150 mi., contain not more than 0·25 g. of
iron, and be about 2N in sulphuric acid. The cold ferric iron
solution is passed through the reductor, using gentle suction,
at a rate not ~xceeding 75 to 100 ml. per minute. As soon
as the reservoir is nearly emptied of the solution, 100 rol. of
2·5 per cent sulphuric acid is p1lssed through in two portions,
followed by 100 to 150 ml. of water. The last washing is
necessary in order to wash out all of the reduced compound
and also the acid, which would otherwise cause unnecessary
consumption of the zinc. Disconnect the flask from the
reductor, wash the end of the delivery tube, and titrate
immediately with standard O·lN potassium permanganate.
. Carry out a blank determination, preferably before passing
the iron solution through the reductor, by running the same
volumes of acid and water through the apparatus as are used
348 Quantitat£ve Inorganic Analysis
in the actual determination. This should not amount to
more than about 0·1 mi. of O·lN permanganate, and should
be deducted from the volume of permanganate'solution used
in the subsequent titra~ion.
It must be emphasised that if hydrochloric acid has been
employed in th.e original solution of the iron-bearing material,
the volume should 'be reduced to ca. 25 ml. and then diluted
to ca. 150 ml. with 5 per cent sulphuric acid. The estimation
is carried out as'detailed above, but 25 ml. of the Zimmermann-
Reinhardt solution must be added before titration with the
standard potasshm'l. permanganate solution. A better method
is to expel the hydrochloric acid by careful evaporation with
an excess of sulphuric acid before the solution is reduced; the
addition of the Zimmermann-Reinhardt solution is then
unnecessary, and the end pbint is more stable.
Applications and limitations of the Jones reductor.
1. Solutions containing 1 to 10 per cent by volume of sul-
phuric acid or 3 to 15 per cent by volume of concentrated
hydrochloric acid can be used in the reduct or. If permanganate
solution is to be used later, hydrochloric acid is best avoided.
2. Nitric acid must be absent, for this'is reduced to hy-
droxylamine and other comp,ounds which react )\'ith perman-
ganate. If nitric acid is present, evaporate the1solution just
to dryness, wash the sides of the vessel with about 3 ml. of
water, carefully add 3 to 4 'illl. of concentrated sulphuric
acid, and evaporate until fumes of the latter are evolved.
Repeat this operation twice to ensure complete removal of
the nitric acid, dilute to 100 ml. with water, add 5 ml. of
concentrated sulphuric acid, and proceed with the reduction.
3. Organic matter (acetates, etc.) must be absent. It is
removed by heating to fumes of sulphuric acid in a covered
beaker (best with a cover of the Fisher '.. spe~dyvap" type),
then carefully adding drops of a saturated solution of potas-
sium permanganate until a permanent colour is obtained, and
finally continuing the fuming for a few minutes.
4. Solutions containing compounds of copper, tin, arsenic,
antimony and other reducible metals must never be used.
These must be removed before the reduction by treatment
with hydrogen sulphide.
o. Other elements which are reduced, in the reductor to a
definite valency stage are titanium to Ti+++, chromium to
Cr++, molybdenum to MO+H and vanadium to V++. Uranium
is reduced to slightly below the quadrivalent state, but by
.bubbling a stream of air through the solution in the filter
_ flask for a few minutes, the dirty dark green colour changes
to the bright apple-green colour characteristic of pure uranous
salts U++++. Tungsten and niobium are reduced, but not to
any definite valency state.
With the exception of ferrous iron and quadrivalent uran-
ium, the reduced solutions are extremely unstable and readily
re-oxidise upon exposure to air. These are best caught in a
five-fold excess of a solution of ferric alum (150 g. of A.R. iron
alum "and 150 mI .. of concentrated sulphuric acid per litre;
approximately 0'3N with respect to iron) contained in the
filter flask. The ferrous iron formed is then titrated with a
standard solution of a suitable oxidising agent. Titanium
and chromium are completely oxidised and lioerate .an
equivalent amount pf ferrous sulphate; molybdenum is re-
oxidised to the quinquevalent (red) stage, which is fairly
stable in air, and complete oxidation is effected by the per-
manganate, but the net resul~is the same, viz., MOIII ~ MOVI ;
vanadium is re-oxidised to the quadrivalent condition, which
is stable in air, and the final oxidation is completed by slQW
titration with potassium permanganate solution.
D. Reduction with sulphurous acid. The solution must be
feebly acid and fairly dilute, say, 500 ml. for 0'5 g. of iron.
If the concentration of the acid exceeds 5N, sulphurous acid
will oxidise ferrous solutions. Hydrochloric acid-chloride
solutions are reduced more rapidly than sulphuric acid-
sulphate solutions. Either sulphur dioxide from a siphon of
the liquid gas or freshly prepared sulphurous acid solution or
ammonium bisulphite solution may be used. The operation
is best carried out in an all-glass special wash bottle or, if tliis
is not available, in a flask fitted with a rubber stopper carrying
two " wash bottle" tubes.
Treat the hydrochloric acid or SUlphuric acid solution of the
iron slowly and with constant shaking with dilute ammonia
solution until a faint permanent precipitate is obtained,
Dilute to about 100 mI., pass sulphur dioxide through the
solution for 2 to 3 minutes, and then gradually heat to boiling,
still continuing the passage of the gas. When the solution is
colourless (15 to 30 minutes), replace the sulphur, dioxide by
a stream of washed carbon dioxide (from a Kipp's apparatus
or cylinder), and boil vigorously until all the sulpJ:lur dioxide
is expelled (20 to 30 minutes) as shown by passing the escaping
gas for 30' seconds through dilute sulphuric acid containing
.
2 drops of O·IN permanganate. Allow. the solution to cool
in a stream of carbbn dioxide, add more acid, and titrate with
a standard solution of a suitable oxidising agent.
A simpler method is to place the acidified iron solution in a
conical flask, add dilute ammonia solution slowly until a
faint permanent precipitate is obtai~d, and then add either
25 ml. of a freshly prepared saturated solution of sulphur
dioxide or excess of freshly prepared ammonium bisulphite
solution: in the latter case, boiling water followed by a little
dilute sulphuric acid is added. A small funnel is placed in
the mouth of the flask, and the mixture boiled for 30 minutes.
All the sulphur dioxide will then have been expelled. Cool
the solution in an atmosphere of carbon dioxide, add 10 ·ml.
of dilute sulphuric acid (1 : 6), and titrate at once with a
standard solution of the oxidising agent. I
Members of the hydrogen sulphide group of metals must be

absent. If present, they .must be removed first.
E. Reduction with hydrogen sUIphide. This method is not
frequently employed. The best experimental conditions are
as follows. About 100 ml; of a cold 2·5 per cent sulphuric
acid solution of the iron compound is treated with a current
of washed hydrogen sulphide for 30 minutes, and then for a
further 15 minutes during which the solution\ is heated to
boiling. 15 Ml. of sulphuric acid (1 : I) are aaded, and the
solution is boiled down to about 50 ml. during 30 to 60 minutes
whilst a stream of oxygen-free carbon dioxide is passed
through. The solution is allowed to cool in a stream of the
gas, diluted to 200 ml. with distilled water, and titrated with
standard permanganate solution. The precipitated sulphur
is coagulated during the concentration, and usually need not
be removed before titration. The results are slightly high.
F. Reduction with zinc and sulphuric acid. This method
has been largely replaced by that employing amalgamated
zinc, but is given here for the sake of. completeness. The
solution to be reduced should be dilute (e.g., not more than
0·15 g. of iron pe); 200 ml.: for this quantity about 10 g. of
zinc and 25 ml. of sulphuric acid are necessary). Place 150
ml. of the iron solution in a 350 ml. conical or flat-bottomed
flask, add cautiously 20 ml. of concentrated sulphuric acid
and 10 g. of A.R. zinc. The reaction will usually be slow,
and is accelerated by the addition of two to three drops of
copper sulphate solution. Warm very gently and allow the
reaction to continue until the solution appears quite colourless
(or pale green): a short funnel should be placed in the neck
of the flask to prevent loss by spray. Test for the presence
of ferric salt by removing a drop of the solution on the end
of a thin glass rod, or at the end of a capillary tube, and
bringing it into contact with a drop of potassium or ammon-
ium thiocyanate solution upon a porcelain "spot" plate.
No red colour should appear, or at
most only a faint pink tint, if all the
ferric iron has been reduced; otherwise
a red colouration is formed. When all
the iron has been reduced, close the
flask by means of a one-holed rubber
stopper carrying a Bunsen valve (Fig.
3-6), and boil the mixture gently until
all the zinc has dissolved. (A Bunsen
valve consists of a piece of glass tubing I
carrying a short length ofrubbertubing, - FIG. 3-6.

which is closed at the upper end by a short glass rod. The rubber
tube has a vertical slit, about 1 to 2 cm. long, forming the
valve; this allows gas or vapour to pass out, but is closed by
~tmospheric pressure when the evolution ceases, thus pre-
venting the entrance of air.) When the zinc has completely
dissolved, allow the solution to cool, remove the stopper,
and titrate the contents of the flask with standard perman-
ganate solution.
For reduction with liquid zinc amalgam, see Section m, 90.
m, 52. Determination of iron in ferric ammonium. sq1-
phate.-In order to obtain experience in the various methods
of ·reduction described in the preceding Section, the student is
recommended to estimate the percentage of iron in. ferric
alum, (NH'.)2S04,Fe2(S04)a,24H20. It is best to employ the
A.R. material, as some check can be made by calculation upon
the accuracy of the result obtained.
Weigh out accurately about 24 g. of A.R. ferric alum, trans-
fer it to a 500 ml. volumetric flask, dissolve it in water con-
taining about 100 ml. of dilute sulphur~c acid (2N; ca. 5 per
cent), and make up to the mark with water. Shake well.
Reduce the iron in 25_ml. portions (i) by the stannous chloride
method, (ii) by the sulphurous acid method, and (iii) by die
Jones reduct or, if available. In all cases the final solution
should be titrated against standard O·lN potassium per-
manganate. Individual determinations should agree within
0·1 ml. of the permanganate solution.
352 . Quantitative Inorganic Analysis
Calculate the percentage of iron in the compound for each
method of reduction. .
1 Ml. N ltMn0 4 == 0·05884 g. Fe.
The yolume of permanganate solution used is multiplied by
its normality; this gives the volume of N solution to which it
is equivalent. If the latter volume is multiplied by the milli-
equivalent of iron (0'05884 g.), the weight of iron in the volume
of solution titrated is obtained, from which the percentage of
iron is readily calculated.
m,53. Determination of the total iron in an iron ore.-For
practice, the student may employ either spathic iron ore
(largely ferrous carbonate)* or the iron ore supplied by the
instructor. Weigh out accurately about 2 g. of the finely-
ground ore and dissolve it in 100 ml. of dilute hydrochloric
acid (1 : 1) in a conical flask carrying a, short funnel in the
neck. Warm gently, and continue the heating until the
residue is free from coloured material. Rinse down the funnel
and the neck of the flask 'with distilled water, cool, and filter
the liquid through a hardened quantitative filter paper into a
250 m!. volumetric flask. Wash with very dilute hydrochloric
acid and allow the washings to'"pass into the flask. Remove
the fu~mel and make up to the mark. Shake well. Remove
50 ml. of the liquid by means of a pipette to. a conical flask,
heat to boiling, and reduce the ferric salt by the stannous
chloride method (Section m, 51 A). Pour the reduced
solution into 400 m!. of water containing 25 zpl. of the
Zimmermann-Reinhardt solution (Section m, 46), and titrate
slowly with standard O·IN potassium permanganate to the
first pink colour which is stable for 15 seconds. Repeat the
titration with two other 50 ml. portions of the soh,ltion.
Calculate the percentage of iron tn'the in;m ore:
ill, 54. Determination of calcium in calcium coJl'bonate.-
Discussion. Metals, such as calcium, copper, le&.d and zinc,
which give slightly soluble oxalates may be determined by
dissolving the washed precipitate in dilute sulphuric acid and
titrating with standard permanganate solution. The method
is widely used for tl).e estimation of calcium. The calcium is
precipitated as the oxalate by the addition of ammonium
oxalate solution to a dilute hydrochloric acid solution of the
element, followed by neutralisation of the acid with dilute
• Suitable iron ores are Ridsdale's .. Iron Ore, No, 17" (one of the Analysed
Samples for Students) and" Iron Orll • A ... (a British Chemical Standard)
of the Bureau of Analysed Samples.
ammonia solution. The washed precipitate. is dissolved in
dilute sulphuric acid, and the liberated oxalic acid titrated
with standard permanganate solution.
CaC 2 0 4 + H 2S0 4 = CaS0 4 + H 2 C2 0 4 ;
2KMnO, + 3H 2S0 4 + 5H 2 C2 0 4
= K 2 S0 4 + 2MnS0 4 + 10C0 2 + 8H 20.
·It must be pointed out that the above method is not quite
satisfactory for the determination of lead, because of the
formation of a coating of lead sulphate over the lead oxalate.
In this case the following procedure, which can be applied to
all the elements given above, should be used. The oxalate
of the metal is precipitated by a known excess of oxalic acid
or ammonium oxalate solution in the presence of ammonia
solution (details are given below), the precipitate is filtered and
well washed, and the excess of oxalate in the filtrate and
washings determined by titration with standard potassium
permanganate solution in the usual manner.
Procedure. Weigh out accurately 0·15 to 0,2 g. of calcium
carbonate, preferably of A.R. grade, into a 400 ml. beaker.
Add 20 m!. of water and cover the beaker with a clock glass.
Add 10 mI. of dilute hydrochloric acid (1 : 1) and warm until
the solid has dissolved. Rinse the sides of the vessel and the
clock glass, dilute to 200 ml., and add a few drops of methyl
red indicator. Heat the solution to boiling. and (j.dd a clear
solution of 1·5 g. of ammonium oxalate dissolved in 25 ml. of
water slowly and with constant stirring; some calcium oxalate
may be precipitated in the acid solution owing to the large
excess used. To the hot solution (70° to 80°C) add dilute
ammonia solution (1 : 1) drop wise and with stirring from a
burette until the liquid is neutral or faintly alkaline (colour
change from red to yellow). Allow the solution to stand for
an hour in a warm place. Decant the clear supernatant
liquid through a What man No. 40 or No. 540 filter paper;
test the filtrate for the presence of calcium with ammonium
oxalate. Transfer the precipitate quantitatively to the filter
(Section n, 28), and wash the precipitate on the filter re-
peatedly with small quantities of cold water until the wash-
ings are oxalate and chloride-free. Pierce a hole in the filter
with a pointed glass rod, and wash the bulk of the precipitate
through the funnel into a conical flask with hot water. Treat
the filter with small quantities of dilute sulphuric acid
(1 : 8), and again wash into the flask. Finally wash the filter
paper thoroughly with hot water. When the precipitate
N
has completely.dissolved (add more dilute sulphuric acid, if
necessary), dilute the solution to about 200 ml., and titrate
with standard O·lN potassium permanganate (for details, see
Section m, 48, Procedure B).
The calcium oxalate precipitate may also be filtered through
a sintered glass or porous porcelai!l crucible. The precipitate
is then washed with dilute sulphuric acid (5 per cent) until
dissolved. The sulphuric acid is added in small portions and
is allowed to remain in contact with the precipitate before
being drawn into the suction flask. The solution in the flask
is ·diluted and titrated with standard O'IN permanganate.
Repeat the determination with two other samples of the
calcium carbonate. -
Calculate the percentage of calcium in the calcium car-
bonate.
I Ml. N KMnO, == 0·02004 g. Ca == 0·02804 g. CaO.
m, 55. Analysis of hydrogen peroxide.-Discussion.
Hydrogen peroxide is marketed in four concentrations,
"10," "20," "40" and "100 volume" concentrations.
The" 10 volume" solution refers to the fact that the solution,
when fully decomposed by heat, yields 10 times its volume of
oxygen measured at O°C and 760 mm.
2H20 2 = 2H zO + 02'
2 X 34'OZ g. 22'4 litres (at N:T.P.)
A " 10 volume" preparation should contain about 3 per

cent H 20 2 • The following reaction occurs when potassium
permanganate solution is added to hydrogen peroxide solu-
tion acidified with dilute sulphuric acid:
2KMnO, + 3H 2 SO, + 5H20 2
= K 2SO, + 2MnSO, + 8H:aO +502'
This forms the basis of the following method of analysis.
Procedure. Transfer 25 ml. of the" 10 volume" solution
by means of a burette (1) to a 250 ml. measuring flask, and
dilute with water to the mark. Shake thoroughly. Transfer
25 mL of this solution to a conical flask, dilute with 200 ml.
of water, add 20 ml. of dilute sulphuric acid (I : 5), and titrate
with standard O'IN potassium permanganate to the first
per~anent, faint pink colour. Repeat the titration; two
consecutive determinations should agree within 0·1 ml.
Calculate (i) the weight of hydrogen peroxide per litre of
the original solution and (ii) the "volume strength," i.e.,
the number of ml. of oxygen at N.T.P. that can be obtained
from 1 ml. of the original solution.
I Ml. N KMnO, :::; 0·01701 g. HzOz.
Note. 1. If the percentage is required by weight, the specific.
gravity of the original solution may be taken as unity. A better
procedure is to weigh out the original sample in a weighing bottle.
,
Ill, 56. Analysis of sodium peroxide.-In this analysis,
precautions must be taken to prevent loss of active oxygen.
Prepare a cold mixture of 100 mI. of water, 5 ml. of concen-
trated sulphuric acid, and 5 g. of pure boric acid.* Transfer
gradually about 0·5 g. of the sodium peroxide from a weighing
. bottle to the mixture, which is kept vigorously stirred. Make
up the solution to 250 m!. in a measuring flask, and shake well.
Titrate 50 ml. portions of this solution with standard O'IN
potassium permanganate.
Calculate the percentage purity of the sample.
1 Ml. N KMn04 == 0'07800 g. Na2 0 2 •
Note. Barium peroxide may be evaluated by dissolving a weighed
quimtity (say I g.) in 200 m!. of N hydrochloric acid, and making up
to 250 m!. in a volumetric flask. 50 M!. portions are titrated with
standard 0, IN potassium permanganate. .
I Ml. N KMnO, = 0·08468 g. BaO z'
The use of the Zimmermann and Reinhardt's solution is nof
necessary in this case (compare Section m, 46).
m,57. Determination of manganese dioxide in pyrolusite.-
Discussion. Manganese dioxide occurs in nature as the mineral
pyrolusite. For many purposes, a knowledge of the percen-
tage of Mn02 is required. This may be determined by treat-
ment with an excess of an acidified solution of a reducing
agent, such as ferrous sulphate, sodium oxalate or arsenious
oxide.
MnOz+ 2FeS0 4 + 2H 2S0 4 = Fe 2(S04)a + MnS0 4 + 2H 20.
Mn02 + H2'C20~'+ H 2S0 4 = MnSO. + 2C0 2 + 2H 2 0.
2MnOg + As 20 a + 2H 2S0 4 = 2MnS0 4 + As 20 S + 2H 2 0.
is
The excess of reducillg agent determined by titration with
standard permanganate solution. The disadvantage of using
ferrous sulphate solution is that the operation must be carried
out in an atmosphere free from air (e.g., carbon dioxide) in
order to prevent the aerial oxidation of the ferrous salt.
Both sodium oxalate and arsenious oxide, which are primary-
• This is added in order that the relatively stable" perboric acid" may be
formed upon the addition of the sodium peroxide.
356 Quantitative Ino'rganic Analysis
standards and are stable in air, are more convenient. Arseni-
ous oxide is somewhat more trustworthy in this estimation
than is sodium oxalate, because oxalic acid decomposes very
slowly at high temperatures into carbon monoxide and carbon
dioxide, the decomposition being catalysed by manganous
salts; the extent of decomposition under ordinary circum-
stances is, however, very small. Both procedures will be
described.
Procedure A (arsenious oxide method). Dry the finely-
powder~d sample of pyrolusite at 120°C to constant weight.
Weigh out accurately from a weighing bottle about 0·2 g. of
the sample into a 250 ml. conical flask, add 50 m!. of standard _
O'IN arsenious oxide (Section m.
48A. Note 1) and 10 mL of
concentrated sulphuric acid. Place a short funnel in the
mouth of the flask, and boil until the pyrolusite has decom-
posed completely; no brown or black particles should then
be present. Cool the solution, add 1 drop of 0·0025M
potassium iodid~ solution, and titrate the excess of arsenious
oxide with standard O·IN potassium permanganate. Repeat
the determination with two other samples of the solid.
Calculate the percentage of MnOz in tne pyrolusite from
the amount of arsenious oxide consumed in the ~itiation.
1 MIl N KMn0 4 =::= 1 m!. N AszO a == {)·04346'g. MnO z.
Procedure B (sodium oxalate method). Weigh out accur-
ately about 0·2 g. of the finely-powdered, dry pyrolusite into a
conical flask, add 50 m!. of standard O·IN sodium oxalate
(Section m, 48B. Note 1), add 50 m!. of 4N sulphuric aCid
(ca. 10 per cent), and place a short funnel in the mouth of the
flask. Boil the mixture gently until no black particles re-
main. Allow to cool, and titrate the excess of oxalate with
standard O·IN potassium permanganate as detailed in Section
m, 48, P,'ocedure B. Repeat the determination with two
other samples of similar weight. '.
Calculate the amount of sodium oxalate c;onsumed in the
reaction, and from this the percentage of MnOz in the pyro-
lusite.
I Ml. N KMnO, = 1 m!. N NazCp, == 0·04346~. MnO,.
Note on the analysis of red lead. A weighed quantity (about
0·3 g.) of red lead is heated with about 30 ro!. of 2N nitric acid:
The follo'Wing reaction occurs:
PbaO, + 4HN0 3 = PbOa + 2Pb(N03)2 + 2HaO·
When the reaction is over, 50 m!. of O'lN sodium oxalate is added,
and the mixture boiled until the solution is almost clear and colour-
less. The excess of the oxalate is then titrated with standard 0·1 N
potassium perm'anganatc.
m, 58. Determination of nitntes.-Discussion. Nitrites
react in warm acid solution (ca. 40 q with permanganate
0
solution in accordance with the equation:

2KMnO, + 3H zSO, + 5HN0 2
. = KzSO, + 2MnSO, + 5HNO a+ 3H 2 0.
If a solution of a nitrite is titrated in the ordinary way with
potassium permanganate, poor results are obtained, because
the nitrite solution has first to be acidified with dilute sul-
phuric acid. Nitr:ous acid is liberated, which being volatile
and unstable, is partially lost. If, however, a measured
volume of standard potassium permanganate solution, acidi-
fied with ~lilute sulphuric acid, is treated with th.e nitrite
solution, added from a burette, until the'permanganate is just
decolourised, results accu'rate to 0·5 to 1 per cent may be
obtained. ,This is due to the fact that nitrous acid does not
react instantaneously with the permanganate: More accurate
rt:sults may be secured by adding the nitrite to an acidified
'solution in which permanganate is present in excess (the tip
of the pipette containing the nitrite solution should be below
the surface of the liquid during the addition), and back-
titrating the potassium permanganate with a solution· of
ferrous ammonium sulphate, which has recently been compared
with the permanganate solution.
the purity of commercial potassium nitrite.
Procedure. Weigh out accurately about 1·1 g. of commercial
potassium nitrite, dissolve it in cold water, and dilute to 250
m!. in a volumetric flask. Shake well. Measure out 25 m!.
of standard O'IN potassium permanganate into a 500 m!.
flask, add 300 mi. of 0'75N sulphuric acid, and heat to 40 e. 0
Place the nitrite solution in the burette, and add.it slowly and
with constant stirring until the permanganate solution is just
decolouI1sed. Better results are attained by allowing the tip
of the burette to dip under tHe surface of the diluted perman-
ganate solution. Towards the end the reaction is sluggish, so
that the nitrite solution must be added very slowly .. Repeat
the titration with two other 25 ml. portions of the perman-
ganate solution.
Calculate the percentage of KN0 2 in the commercial sample.
1 Ml. N KMn04 == 0·02351 g. HN0 2 == 0·04256 g. KN0 2 •
m.59. Determination of persulphate~.-Discussion. Alkali
persulphates can readily be evaluated by adding to their
solutions a known excess of an acidified ferrous salt solution,
and determining the excess of ferrous iron by titration with
standard potassium permanganate solution.
K 2S 2 0 S + 2FeS0 4 + H 2S0 4 = Fe 2 (S04)a + 2KHS0 4 •
For practice, the student may determine the purity of
potassium persulphate, preferably of A.R. quality.
Procedttre. Prepare an approximately O·IN solution of
ferrous ammonium sulphate as described in Section m. 48.
Procedure D, and titrate this against standard O'IN potassium
permanganate. Weigh out accurately about 0·3 g. of potas-
~ium persulphate into a conical flask fitted with a Bunsen
valve (Fig. 3-6). Remove the rubber stopper carrying the
valve, displace the air by carbon dioxide, add 50 ml. of the
acidified ca. O·IN ferrous solution, followed by 150 to 200
ml. of boiling water so as to raise the temperature tq at least
80°C. Close the flask and mix the contents thoroughly.
When the solid has dissolved, cool the solution by immersion
in cold water, and titrate the exc~ss of ferrovs irbn against
standard O'IN potassium permang'anate. Repeat the titra-
tion with two further quantities of the potassium persulphate .
.From the difference behyeen the volume of O'IN perman-
ganate required to oxidise 50 ml. of the ferrous solution and
that required to. oxidise the ferrous salt remaining after the
addition of the persulphate, calculate the percentage purity
of the sample.
1 Ml. N KMn04 ='0·1352 g. K 2S 20 S '
m. 60. Determination of manganese in steel (bismuthate
method).-Discussion. Sodium bismuthate of commerce is a
brown powder of somewhat indefinite composition; the
oxidising component is usually taken as NaBiO a. The A.R.
product contains about 85 per cent of NaBiO s' Manganous
salts are oxidised to permanganic acid by excess of sodium
bismuthate in the presence of nitric acid:
2Mn(NO a)2 + 5NaBiO s + 16HNO a
= 2HMn04 + 5Bi(NO a)s + 5NaNO a ~ 7H zO.
The permC;l,nganic acid is very stable in a cold solution con-
taining 20 to 40 per cent of nitric acid. If the steel is dis-
V olitmet'ric Analysis 359
solved in nitric acid and the resultant cold solution treated
with excess of sodium bismuthate, permanganic acid is formed.
The excess of bismufl).ate is removed by filtration, and a
measured excess of standard ferrous solution added to reduce
the permanganate to a manganous salt :
MnO,- + 5Fe++ + 8H+ = Mn++ + 5Fe+++ + 4H 20.
The excess of ferrous iron is then determined by titration
with standard potassium permanganate solution.
The solution should be free from cobalt, chromium, and
chloride.
Procedure. Prepare the following solutions:
O'03N Potassium permanganate. Dissolve 1 g. of A.R.
potassium permanganate in 1 litre of water (Section m, 47).
O'03N Ferrous ammonium sulphate. Dissolve 12 g. of A.R.
ferrous ammonium sulphate in 1 litre of 5 per cent sulphuric
acid (Section m, 48, Procedure D).
3 Per cent nitric acid free from nitrous acid. Boil concen-
trated nitric acid (d 1·42) for a short time (fume cupboard).
Dilute in the proportion of 3 ml. of the concentrated acid to
100 m!. of water. (One m!. of nitric acid, d 1'42, contains
nearly 1 g. of HNO a. The above is therefore a 3 per cent
solution both by volume and by weight.)
Standardise the permanganate solution against A.R.
sodium oxalate either by weighing out accurately about 0·1 g.
of the dry oxalate and proceeding as in Section m, 48 (Pro-
cedure B) or by preparing an approximately 0'03N solution
(0·5 g. per 250 ml.) and following the experimental procedure
of Section m, 48, Procedure B, Note 1. Alternatively, the
arsenious oxide method (Section In, 40, Procedure A) may be
used.
Weigh out accu.rately aoout 1 g. of the sample containing
0·2 to 1 per cent of manganese, * and dissolve it in 50 m!. of
dilute nitric acid (1 : 3) in a 250 m!. conical flasJs. Filter, if
necessary. through a sintered glass crucible. Boil for 5
minutes, cool, add 0·5 g. of A.R. sodium bismuthate. and boil
again for 5 minutes. A pink colour due to permanganic acid
and/or a precipitate of manganese dioxide shol1ld appear;
if neither a colouration nor ~ precipitate is produced, add a
further 0·5 g. of sodium bismuthate and boil again for 5
minute;s (1). Then add dropwise a concentrated solution of
• Either Ridsdale's .. Medium Carbon Steel, No.2" (one of_the Analysed
Samples for Students) or the Bureau of Analysed Samples .. Carbon Steel
, E ' or ' R ' " (a British Chemical Standard) is suitable.
freshly prepared sulphurous acid or of sodium sulphite until
the solution clears. Boil until all oxides of nitrogen are
expelled (2). Cool the solution to f!.bout 15°C, add A.R.
sodium bismuthate until no further appreciable alteration in
colour ensues (about 0'25 g. NaBiO a for each 0·01 g. of Mn),
add a further ()'5 g. excess, and stir the mixture for 2 to 3
minutes. Add 50 ml. of the prepared 3 per cent nitric acid,
and filter through a Gooch (3) or sintered glass crucible into a
350 ml. conical flask. Wash the residue well with the 3 per
cent nitric acid until free from permanganic acid; this will
generally require 50 to 100 ml., and the approximate volume
should be noted. Run in slowly the ferrous ammonium
sulphate solution from a burette until a slight excess is present
(note the exact volume) as indicated by the disappearance of
the permanganate colour (4). Titrate immediately with the
o·OaN potassium permanganate to the first faint pink coloura-
tion (5). The addition of 3 ml. of syrupy phosphoric acid
improves the end point.
Standardise the ferrous ammonium foolution by running a
blank in the following manner. Into a 350 m!. Erlenmeyer
flask place 50 m!. of 1 : 3 nitric acid, add a little sodium
bismuthate, dilute with 50 to 100 m!. (use the volume of the
wash liquid) of 3 per cent nitric acid, add an equal volume of
the ferrous solution as was used in the determination, and
titrate against the standard potassium permanganate solu-
tion. The difference between the titrations represents the
permanganate equivalent of the permanganic acid formed in
the determination.
Calculate the percentage of manganese in the sample of
steel.
1 MI. N KMn04 ==: 0·01099 g. Mn.
Notes. 1. This preliminary oxidation with sodium bismuth ate
is to destroy any reducing substances, including carbonaceous
matter, pres~nt which would subsequently slowly react with the
permanganate.
2. All oxides of nitrogen must be removed, since these react with
the permanganate. The reduction will also dissolve any manganese
compounds that may. have separated from the solution through the
decomposition of the permanganic acid owing to the vigorous boiling.
3. The Gooch crucible containing asbestos must be freshly ignited
before use.
4. The ferrous solution must be added as soon as the filtration
and washing are complete, because the permanganic acid slowly
decomposes.
5. This titration must be carried 'out immediately the ferrous
solution has been added, because the ferrous solution reacts slowly
with the nitric acid.
m, 61. Determination of formates and of formic acid.-
Discussion. Potassium permanganate reacts slowly with
formic a:cid in cold acid solutions; in hot solutions, loss by
volatilisation occurs. Formates in the presence of an alkali
carbonate, e.g., sodium or potassium carbonate, react quan-
titatively with potassium permanga·nate ;
2KMnO, + 3H.COOK = KHCO a +_2K 2CO a+ 2Mn0 2 + H 20.
This is one of the rare examples in which potassium perman-
ganate has an equivalent weight of 1/3 of the formula weight
ox; 1/3 mol:
MnO,- + 2H 20 + 3€ = Mn02 + 40H-,
or reacts in accordance with " the hypothetical formula
K 2Mn 20 S = K 20,2Mn0 2 , 30. In this method, the formic acid
or formate is treated with an excess of sodium carbonate, and
the standard potassium permanganate solution allowed to run
into the hot formate solution until the clear liquid above the
precipitate is just coloured pink. Some difficulty may be
experienced in detecting the exact end point in the presence
of the brown precipitate. The following procedure is there-
fore recommended. After the solution has acquired a distinct
pink colour, it is strongly acidified with dilute sulphuric acid.
a known excess of approximately O'IN sodium oxalate
solution added, and the mixture warmed until the precipitate
has dissolyed. The excess of oxalate is then titrated yvith
standard potassium permanganate solution.
In the modified procedure, the final end point is produced in
acid solution. The final products of a chemical reaction are
independent of the stages by which that reaction has been
brought to completion, hence the permanganate reacts in
accordance with the ionic equation:
Mn0 4- + 8H+ + 5€ = Mn++ + 4H 2 0,
(or as K 2 0,2MnO,50) and its equivalent is 1/5 mol. The
oxidation of the formic acid is essentially:
H.eOOR + 0 = H 2 0 + CO 2 ,
hence 1 m!. N KMnO, == 0·02301 g. HCOOR.
The method is unreliable in the presence of much chloride.
For practice, the stud~nt may determine the purity of
commercial sodium formate.
N*
362 Quantitative Inorga11;ic Analysis
Procedure. Weigh out accurately about 1,7 g. of sodium
formate and dissolve it in 500 ml. of water in a volumetric
flask. Shake well. Transfer 25 ml. of this soiution into a
conical flask, add excess ·of sodium carbonate solution,
followed by standard O·IN potassium permanganate from a
burette to the hot solution uhtil the liquid has a distinct pink
colour. Acidify strongly with dilute sulphuric acid, add 25
m!. of O·IN sodium oxalate, and warm until the precipitate
has dissolved and the permanganate colour has disappeared.
Titrate the excess of oxalate against the standard perman-
ganate solution. Titrate 25 m!. of the sodium oxalate solu-
tion with the permanganate solution; the difference between
this result and the .total permanganate solution used gives the
quantity of permanganate required to oxidise the formate.
Repeat the determination with two other 25 ml. portions of
tfie formate solution. .(
Calculate the percentage purity of the sample of sodium
formate.
m, 62. Determination of selenium.-Method A. The
selenious acid or selenite corresponding to about 0-1 g. of
selenium is dissolved in 25 ml. of 40 per cent sulphuric aCid,
and diluted to 150 ml. Twelve grams of sodium phosphate
(or phosphoric acid) is added (to prevent formatiop of man-
ganese dioxide), followed by a considerable exces~, at least
10 ml., of standard O·IN potassium permang!lnate. After
30 minutes, the residual potassium permanganate is deter-
mined by the addition of a slight excess of O·IN ferrous ammon-
ium sulphate and back titration with standard O'IN perman-
ganate.
1 MI. N KMn0 4 == 0·03948 g . ..se.
Method B.* The reaction:
H 2SeO a + 4HI = Se + 212 + 3H 20
is quantitative if carbon disulphide is present to prevent the
liberated' iodine from being adsorbed on the selenium. 100
Ml. of an aqueous solution containing 0'05 to 0'10 g. of selen-
ium as selenious acid are placed in a 250 m!. glass-stoppered
bottle containing 10 ml. of 25 per cent hydrochloric acid and
20 ml. of carbon disulphide. Ten ml. of 10 per cent potas-
sium iodide solution are added in a fine stream whilst the
liquid is rotated. Shaking is continued for 1 minute, and the
'" This method should be given under Iodimetry and Iodometry, but is
described here in order to have both methods for the determination of selenium
together.
liberated iodine is then titrated with standard O'lN sodium
thiosulphate with shaking, starch being added as indicator.
1 Ml. N Na 2S 2 0 a '= 0'01974 g. Se.
Note. A mixture of selenious and tellirrous oxides or acids can be
estimated as follows. By the use of 0 ·IN potassil,lm permanganate
both oxides or acids are oxidised to the sexavalent condition. With
(J·IN potassium dichromate, tellurium alone (see Section m, 80,
Method B), is so oxidised. Hence by carrying out each oxidation
on a different portion of the solution, each element can be deter-
mined.
It is not essential, however, to carry out titrations with two
different samples, for if one portion is titrated with potassium
dichromate solution, the tellurous tellurium will be oxidised; the
titration can then be completed with potassium permanganate
solution. The pink colour appears sharply; at the true end point
the green colour of the chromic salt and the nrst drop of excess of
permanganate solution gives a neutral grey tint. The effect is
similar to that of " screened" methyl orange; the colour change is
green ...... grey ...... red.
OXIDATIONS WITil POTASSIUM DICHROMATE
m, 63. General discussion.-Potassium dichromate is not
so powerful an oxidising agent as potassium permanganate
(compare oxidation potentials in TableXVI,and also Fig. 1-17),
but it has, however, several advantages over the latter
substance. It can be obtained pure, is stable up to its fusion
point and is therefore an excellent primary standard. Stan-
dard solutions of exactly known strength can be prepared by
weighing out the pure dry salt and dissolving it in the proper
volume of water. Furthermore, the aqueous solution are
stable indefinitely. Potassium dichromate is used only in
acid solution, and is reduced rapidly at the ordinary temper;t-
ture to a green chromic salt. It is not reduced by cold
hydrochloric acid, provided the acid concentration does not
exceed 1 or 2N. Dichromate solutions are also less easily
reduced by organiC matter than are those of permanganate.
Potassium dichromate is therefore of particular value in the
determination of iron in iron ores: the ore is usually dissolved
in hydrochloric acid, and the ferric iron reduced to the ferrous
state with stannous chloride solution (see Section m, 51A) :
K 2Cr Z0 7 + 6FeC1 2 + 14HCl = '2KCI + 2CrC1 3 +6FeCI 3 +7H 2 0.
In acid solution, the reduction of potassium dichromate
may be represented as:
Cr 20 7- - + 14H+ + 6E = 2Cr+++ + 7H 2 0,
from which it follows (compare Section 1,23) that the equiva-
lent weight is one sixth of the molecular weight, 294·21/6 or
49·305 g. A 0-lN solution therefore contains 4-9035 g. per
. litre.*
With regard to the determination of the end point in titra-
Hons with potassium dichromate solutions three methods are
available: (i) with an external indicator, (ii) with an internal
indicator, and (iii) potentiometrically_ A general discussion.
of the principles underlying the potentiometric method has
been given in Section 1, 51.
* For elementary students, the equivalent weight may be deduced from the
hypothetical equations:
K.Cr.0 7 + 4H.SO. = K.SO. + Cr.(SO.la + 4H.0 + 30 ;
K.Cr.O, + SHCI = 2KCI + 2CrCI 3 + 4H 2 0 + 30.
From either of these equations, it follows that the equivalent weight is one
sixth of the mol"cular weight (K.Cr20 7/6).
364
I
Volumetric Analysis :l65
The external indicator method was formerly widely used,

but has become obsolete since the introduction of suitable
oxidation-reduction indicators. It will only be described here
for the sake of completeness, and also as an illustration of the
use of an external indicator. Its widest application was in
connexion with the titration of ferrous iron. A small crystal
of pure potassium ferricyanide is repeatedly washed with
distilled water in order to remove the superficial coating of the
ferrocyanide. The washed. crystal is dissolved in distilled
water (:1> 0·1 per cent solution) and drops of the solution,
which should appear almost colourless or pale yellow, are
placed on a white tile with the aid of a stirring rod. The
acidified ferrous solution is titrated against the standard
potassium dichromate solution. A drop of the ferrous
solution is withdrawn from time to ti!Ue with a thin glass rod
ana placed adjacent to a drop of the indicator; the two drops
are caref,?-lly allowed to coalesce.* At first a deep blue colour
is produced where the drops mix, but as the addition of the
dichromate solution is continued, this is replaced by a bluish
green, then green shade, and the titration is completed when
there is no trace of green and no change is apparent in the
mixed drops. After the first titration, the dichromate solu-
tion is run in slowly into the diluted ferrous solution without
testing to within 0·5 ml. of the previously determined end
point, and then added a drop at a time, the liquid being tested
after the addition of each drop.
The first internal indicator to be used for this titration was
a 1 per cent solution of diphenylamine in concentrated
sulphuric acid (J. Knop, 1924). Subsequently a 1 per cent
solution of diphenylbenzidine in concentrated sulphuric acid
and also a 0·2 to 0·3 per cent aqueous solution of sodium
diphenylamine sulphonate were introduced. With these
three indicators, phosphoric acid must be added; the reason
for this is that the phosphoric acid lowers' the oxidation
potential of the ferric-ferrous system by forming a complex
with the ferric irons (for further details, see Section I, 50).
These indicators impart a green colour to the ferrous solution,
... The above procedure may be replaced by the following which •. although
more time-consuming. has the advantage that the potassium fcrricyanide
solution is not exposed to the atmosphere of the laboratory for longer than is
absolutely necessary. One drop of the indicator is set out on a white tile
with a thin glass rod. the latter is washed and then employed to withdraw a
drop of the titrated liquid, and the drops allowed to mix as described. The
process is then repeated-the rod is washed, a drop of the ferricyanide solution
placed on the tile, the rod washed again, a drop of the solution removed. etc.,
until the end point is reached.
,
which deepens to a blue-green shortly before the en~ of the
titration. At the end poin1: an intense purple or blue-violet
colouration is obtained, which remains permanent after
shaking. The addition of phosphoric acid may be avoided by
the use of N-phenylanthranilic acid or of tri-ortho-phenan-
throline ferrous sulphate. The former is comparatively
inexpensive, and is prepared by dissolving 1·07 g. of N-
phenylahthranilic acid in ~O ml. of 5 per cent sodium carbonate
solution and diluting to 1 litre; about 0·5 ml. is used in a
titration. Tri-ortho~phenanthroline ferrous sulphate* is an
expensive indicator, but since only one drop is necessary for
each titration in a volume of 100 to 150 ml. owing to the
intense colouration produced, the cost is in' reality not high.t
With both indicators it is best to work in a volume of approxi-
mately 300 ml. and the solution should be about 2N with
respect to sulphuric acid; satisfactory results are also ob-
tained when the solution is about 2N with respect ~o hydro-
chloric acid. If the dichromate solution is .placed in the
burette, the colour change at the end point is from orange to
pale green for ferroin, and from green to violet red for N-
phenylanthranilic acid. The reverse titration, dichromate
with ferrous solution, is slower b1J.t definite, and the end point
must be approached more slowly.
It has already been mentioned that a standard solution of
potassium dichromate can be prepared by weighing out
accurately the requisite quantity of the pure dry salt and dissolv-
ing it in the appropriate quantity of water. Further stan-
dardisation is unnecessary. Sometimes, however, the pro-
blem of standardising a dichromate solution does arise. The
best method is to use A.R. iron wire of 99·9 per cent purity
(see Section m, 46). Ferrous ammonium sulphate may be
employed in elementary work or in work requiring only moder-
ate accuracy (see Section III, 46).
m, 64. Preparation of O.IN ·potassium dichromate.-A.R.
potassium dichroma.te has a purity of not less than 99·9 per
cent.! Powder fi!1ely about 6 g. of the A.R. product in a
* The abbreviation" ferroin "will be employed in all subsequent references
to this indicator.
t A 4 oz. bottle of the prepared indicator, which is sufficient for about
2,000 titrations, costs abont lOs. or 2$. This is a product of the G. F. Smith
Chemical Co., 867 McKinley Avenue, Columbus, Ohio, U.S.A. It may also
be purchased from the London agents, F. W. Berk and Co. Ltd., 52-54
Leadenhall Street, London, E.C.2, England.
t If a greater purity is required, or if only a pure grade of commercial salt
is available, or if there is some doubt as to the purity of the salt, the following
method of purification should be used. A concentrated solution of the salt
glass-or agate mortar, and heat for 30 to 60 minutes in an air
oven at 140° to 150°C. Allow to cool in a desiccator. Weigh
out accurately about 4·9 g. of the dry potassium dichromate
into a weighing bottle and transfer the salt quantitatively to a
1 litre measuring flask (Section fi, 6A (iii), using a small
funnel to avoid loss. Dissolve the salt in the flask in water
and make up to the mark; shake well. Alternatively, place
a little over 4'9 g. of potassium dichromate in a weighing
bottle, and weigh accurately. Empty the salt into a litre
volumetric flask, and weigh the bottle again. Dissolve the
salt in water, and make up to the mark. Mix thoroughly.
c;alculate the exact normality by d.ividing the actual weight
of the potassium dichromate employed by the theoretical
weight for I litre of normal solution (49'035 g.). An exactly
O'IOOON solution may be prepared: by weighing out 4·904 g.
of the salt and dissolving it in I litre of water in a measuring
flask.
m, 65. Standardisation of potassium dichromate solution
against iron.-For practice in this determination, use the
O-IN potassium dichromate prepared in the previous Section.
Use the method described in Section III, 48, Procedure C,
with 0-2 g., accurately weighed, of A.R. iron wire. Alter-
natively, a flask fitted with a Bunsen valve (Fig. 3-6) may be
employed: the air in the flask is similarly displaced by carbon
dioxide. Titrate the cooled solution immediately against the
dichromate solution, using either sodium diphenylamine
sulphonate or N-phenylanthranilic acid as indicator. If the
former is selected, add 6 to 8 drops of the indicator, followed
by 5 m!. of syrupy phosphoric acid: titrate slowly with the
dichromate solution, stirring well, until the pure gr_een colour
changes to a grey-green. Then add the dichromate solution
dropwise until the first til1ge of blue-violet, which remains
permanent on shaking, appears. If the latter indicator is
selected, add 200 m!. of 2N SUlphuric acid, then 0·5 m!. of the
indicator; add the dichromate solution, with shaking until
the colour changes from green to violet-red. Repeat the
determination with two similar quantities of iron.
Calculate the normality of the potassium dichromate
solution.
I Ml. N K 2Cr Z0 7 == 0·05584 g. Fe.
in hot water is prepared and filtered. The crystals which separate on cooling
are filtered on a sintered glass filter funnel (Section II. 12A) and sucked dry.
The resultant crystals are recrystallised again. The purified crystals are then
dried at 180° to 200° C, ground to a fine powder in a glass or agate mortar, and
again dried at 14U o to 150°C to constant weight.
Quantitative Inorganic Analysis
Note. For wor_k. requiring only moderate accuracy, A.R. ferrous
ammonium sulphate may be used for standardisation. An approxi-
matrly () ·IN solution is prepared by weighing out accurately about
n·g g. of the solid and dissolving it in 250 ml. of 5 per cent sulphuric
acid in a volumetric flask (Section m, 48, Procedure D). This is
SIOle>ly titrated again~t the dichromate solution, using sodium
diphenylamine sulphonate, N-phenylanthranilic acid or ferroin as
indicator.
POTASSIUM DICHROMATE SOLUTIONS.
m, 66. Determination of ferrous iron.-For practice in the
titration, the student should determine the percentage' of
iron in crystallised ferrous sulphate FeS0 4 ,7H 20, preferably
of A. R. grade.
Weigh out accurately about 14. g. of ferrous sulphate
crystals, dissolve in 450 ml. of 5 per-cent sulphuric acid in a
fiOO ml. measuring flask, and make up to the mark with dis-
tilled water. Shake well. Jitrate 25 ml. portions against
the standard O'lN potassium dichromate with sodium
diphenylamine sulphonate (1)* and with N-phenylanthranilic
acid (II) as examples of internal indicators, and with potas-
sium ferricyanide as an external indicator.
Use 8 drops (say 0·4 m!.) of the indicator I, add 200 ml. of
. 2·;; per. cent sulphuric acid, followed by 5 ml. of 85 per cent
phosphoric acid, and titrate slowly and with constant stirring
against the standard dichromate untif the solution assumes a
bluish-green or greyish-blue tint near the end point. Con-
tinue the titration, adding the dichromate solution dropwise
and maintaining an interval of a few seconds between each
drop, until the addition oCone drop causes the formation of an
intense purple or violet-blue colouration, -which remains
permanent after shaking and is unaffected on the further
addition of the dichromate. Carry out 2 or at most 3 titra-
tions; these should agree within 0'1 ml.
Use 0·5 ml. of indicator II. Add about 200 ml. of 2N
SUlphuric acid and then titrate against the O'lN potassium
dichroma.te nntil the colour changes from green to violet-red.
This titration is sharp to within 1 drop. Carry out 2 or 3
'furtller titrations, which should agree within 0·1 ml.
For the determination ~'ith the external indicator, use 25
ml. of the ferrous solution, add 100 m!. of N (ca. 2·5 per cent)
sulphuric acid, and proceed as detailed , in Section In, 63.
* Or with ,liphenylarninc or diphcnylbcllzidine; use:3 drops of the I per
cent solution ill concentrated sulphuric acid.
As in previous instances, carry 'out about three titrations; two
successive values should agree within 0·1 ml.
Compare the mean volumes of the dichromate solution
consumed in the three tit rations with different indicators.
These should not differ by more than 0·1 ml. Calculate the
percentage of iron in the crystals.
I Ml. N K 2Cr20 7 == 0·05584 g. Fe.
m, 67. Determination of' ferric iron (iron in ferric ammon-

ium sulphate).-Weigh out accurately about 12 g. of A.R.
ferric alum (NH4)2S04,Fe2(S04)a,Z4H20, transfer to a 250
ml. volumetric flask, dissolve in water containing about 50
ml. of 2N (ca. 5 per cent) sulphuric acid, and make up to the
mark with water. Shake well.
Of the various methods which can be employed for the
reduction of the ferri<f iron, the stannous chloride method is
very conveni~nt, and, furthermore, illustrates the advantage
of the use of dichromate solution when hydrochloric acid is
present (compare Section m, 46).
Remove 25 ml. of the ferric solution, and reduce it with
stannous chloride solution as described in Section m, 51 A.
Then add 200 m!. of 2'5 per cent sulphuric acid, 5 m!. of
syrupy phosphoric acid, and 0·3 to 0·4 ml. of sodium
diphenylamine sulphonate indicator. * Titrate slowly and
with constant stirring against standard O'IN potassium
dichromate to the first permanent yiolet-blue colouration.
Carry out 2 further titrations; these should p_gree within
0·1 ml.
Calculate the percentage of iron in ferric alum.
m, 68. Determination of the total iron in an iron ore.-
The iron ore of Section m, 53 may be used, if desired. Weigh
out accurately about 2 g. of the iron ore, dissolve it in dilute
hydrochloric acid, and proceed exactly as described in Section
III, 53 to the point where a solution in a 250 m!. volumetric
flask is obtained. Remove 50 m!. of this solution, heat to
boiling, and reduce the ferric iron by the stannous chloride
method (Section m, 51 A). Complete the estimation as
detailed in the previous Section. Carry out three separate
titrations, which should agree within 0'1 mI., and take the
mean of the results.
Calculate the percentage of iro~ in the ore.
• Alternatively 3 drops of diphenylamine or of diphenylbcnzidine indicator
may be used.
370 Quantitative Inorganic A naJysis
m, 69. Determination 01 lerrous and lerric iron in an iron
ore.-For this estimation, spathic iron ore or any other suit-
able iro~ ore may be used. Determine first the total iron by
the method described in the preceding Section.
Determine the ferrous iron as follows. Weigh out accur-
ately about 0·4 g. of the finely-powdered iron ore into a 350
ml. flask fitted with wash-bottle tubes; displace the air by
carbon dioxide from a Kipp's apparatus or from a cylinder of
the gas fitted with a reducing valve. Open the flask momen-
tarily and pour in 30 ml. of I : I-hydrochloric acid. Replace
the rubber stopper and tubes, and warm in a slow stream of
carbon dioxide until the ore has been completely attacked and
no further solution takes place. A whitish residue of silica
remains in most cases. Cool in a stream of carbon dioxide,
wash down the tubes and the neck of the vessel with a little
cold, air-free distilled water, dilute with 200 ml. of 2·5 per cent
sulphuric acid (which has been pr.epared with air-free water),
add 5 ml. of syrupy phosphoric acid and 0·3 to 0·4 ml. of
sodium diphenylamine sulphonate indicator, and titrate
slowly and with constant stirring against standard O'lN
potassium dichromate to the first permanent violet-blue
colouration. Alternatively, 1 or ~ drops of ferroin may be
used as indicator; here the addition of phosphoric acid is
superfluous. Repeat the determination with two other I
similar quantities of· the ore. ,

Calculate the percentages of total iron and ferrous iron in
the ore'. The difference between these two values gives the
percentage of iron present in the ferric state.
m, 70. Determination of chromium in a chromic salt.-
Discussion. Chromic salts are oxidised to dichromates by
boiling with excess of a persulphate solution in the presence
of a little silver nitrate (catalyst), The excess of persulphate
remaining after the oxidation is complete is destroyed by
boiling the solution for a short time. The dichromate content
of the resultant solution is determined by the addition of
excess of a standard ferrous solution and titration of the
excess of the latter with standard O'IN potassium dichromate.
For practice in this determination, the student may
determine the percentage of chromium in A.R. chrome alum,
K 2SO 4, Cr2(SO ,) 3,24H 20.
Procedure. Weigh out accurately about 2·5 g. of chrome
alum, and dissolve it in 50 rhl. of distilled water. Add 20 ml.
of ca. O'IN silver nitrate solution, followed by 50 ml. of a 10
per cent solution ot ammonium or potassium persulphate.
Boil the liquid gently (using a Fisher" speedyvap " beaker
cover) for 20 minutes. Cool, and dilute to 250 ml. in a
volum,etric flask. Remove 50 m!. of the solution with a
pipette, add 50 ml. of O·IN ferrous ammonium sulphate
solution (Section m, 48, Procedure D), 200 mI. of 2N sulphuric
acid and 0·5 m!. of N-phenylanthranilic acid indicator.
Titrate the excess of the ferrous salt with standard 0'IN
potassium dichromate until the colour changes from green to
violet-red.
Standardise the ferrous ammonium sulphate solution
against the O·IN potassium dichromate using N-phenylan-
thranilic acid as indicator (Section m, 66). Calculate the
volume of the ferrous solution which was oxidised by the
dichromate originating from the chrome alum, and from this
the percentage of chromium in the sample.
1 Ml. N Fe++ == 0·01734 g. Cr.
m, 71. Determination of chroniium in chromite.- Discus-
sion. The highly refractory mrneral chromite is brought into
solution by fusion with excess of sodium peroxide:
2FeCr 2 0, + 7Na20 2 = 2NaFe0 2 + 4Na 2Cr0 4 + 2Na 20.
Up'on leaching the melt with water, the sodium chromate
dissolves and the iron is precipitated as ferric hydroxide:
NaFe0 2 +
2H 2 0 = NaOH + Fe(OH)3; .
2Na20 + 4H 2 0 = 4NaOH.
The excess of peroxide is decomposed by boiling the alkaline
solution. The precipitate is filtered off after diluting the
solution; the filtrate is acidified with hydrochloric acid, a
known volume of excess of ferrous ammonium sulphate solution
is added, and the excess of ferrous salt is titrated with standard
potassium dichromate solution.
2Na2Cr0 4 + 2HCl = Na 2Cr 20 7 + 2NaCl + H 20;
Na 2Cr20 7 + 6FeS0 4 + I4HCI
= 2CrC1 3 + 2NaCl + 2Fe 2(SOJs + 2FeC1 3 + 7H 2 0.
Procedure. Weigh out accurately about 0·5 g. of the very
finely powdered ore* into a 30 to 35 m!. nickel, or heavy-
walled porcelain, crucible, add 4 g. of sodium peroxide; and
mix thoroughly by means of a thin glass rod. Remove any
* Ridsdale's " Chrome Iron Ore, No. 27 .. (one of the Analysed Samples for
Students) is suitable for this determination.
powder ~dheri.ng to the rod by stirring about I g. of sodium
pero,xide with it; cover the mixture in the cruci,ble with this
peroxide. Place the lid on the crucible, and gently heat the
covered crucible in the fume cupb9ard over a small flame
until the mass is quite liquid (about 10 minutes); keep fused
for a further 10 minutes at a dull red heat. Allow to cool,
and when a solid crust has formed, add 4 g. more of the
sodium peroxide, and fuse the mixture again at a cherry-red
heat for 10 minutes. Allow the crucible to cool and place it
in a 600 mI. Pyrex or Hysil beaker containing a little distilled
water. Cover the beaker with a clock glass (or with a Fisher
" speedyvap beaker cover), add a little warm water, and,
II
after the violent action has subsided, remove the crucible and
wash it thoroughly, collecting the washings in the same beaker.
Boil the liquid for 30 minutes, keeping the beaker covered
(this decomposes the hydrogen peroxide), add 250 ml. of
boiling water, and allow the precipitate to settle. Filter
through a hardened 15 cm. filter paper or, better, through a
sintered glass crucible, * and wash the residue thoroughly
with- boiling water until free from chromate. (The residue
should be completely soluble in concentrated hydrochloric
acid; no black gritty particles should remain. If this is not
the case, decomposition is not complete, and the determina-
tion must be started afresh.) Evaporate the filtrate to about
200 mI., cool, and add 9N sulphuric acid cautiously until acid.
Cool, and transfer to a 250 m!. measuring flask, and make up
to the mark with distilled water. Shake well. Remove 50
m!. of this solution with a pipette, add 50 m!. of O'IN ferrous
ammonium sulphate (Section m, 48, Procedure D), 200 ml. of
2N sulphuric acid, and 0·5 m!. of N-phenylanthranilic acid
indicator. Titrate with standard O·lN potassium dichromate
until the colour changes from green to violet-red. Alterna-
tively, take 50 ml. of the solution, add 50 ml. of O·lN ferrous
ammonium SUlphate, [J ml. of syrupy phosphoric acid and 0·4
ml. of sodium diphenylamine sulphonate indicator; dilute to
200 ml., and titrate with standard O'IN potassium dichromate
until the green colour changes to violet-bIJle. Repeat until
two titrations agree within 0·1 ml.
The ferrous ammonium sulphate solution must be stan-
dardised by titration against the O'lN potassium dichromate;
the same indicator as was employed in the above determina-
tion must be used. .
• If the latter is used, it is unnecessary to add more than 1.50 m!. of boiling
water; the subsequent concentration is then avoided.
Volumetr£c Analysis 373
Calculate the volume of the ferrous solution which was
oxidised by the dichromate originating from the chromite, and
from this the percentage of chromium in the original ore.
1 Ml. N Fe++ == 1 m!. N K 2Cr 2 0 7
== 0·02534 g. Cr20 3 == 0·01734 g. Cr.
m, 72. Determination of manganese in steel or in mangan-
ese ore (Pattinson's method).-Discussion. When a solution
containing iron, zinc and manganese salts is treated with
bromine water and calcium carbonate, all the manganese is
precipitated in the quadrivalent condition as the manganite
(e.g.,. ZnO,MnO z). The precipitate is dissolved in excess of
ferrous solution, and the excess of the latter titrated with
standard potassium dichromate ~olution.
Mn02 + 2FeSO, + 2H 2SO, = Fe 2 (SO,)a + MnSO, + 2H 2 0.
Procedure. Weigh put 2·5 g. of the steel or 1 g. of the man-
ganese ore* and dissolve it in 50 ml. of concentrated hydro-
chloric acid in a covered beaker. Oxidise with a little nitric
acid and evaporate to a small volume in order to expel the
excess of acid. Transfer to a 250 ml. measuring flask, after
filtering through a si.ntered glass crucible, if necessary. Make
up to the mark with distilled water, and shake thoroughly.
Tra.nsfer 50 ml. of the solution into a litre beaker, add 30
m!. of 3 per cent zinc chloride solution, and then add pure
calcium carbonate in small quantities at a time until the
liquid is almost neutralised, but still remains clear. Add
100 ml. of saturated bromine water, and then more calcium
carbonate with constant stirring until an excess of about 1
gram remains undissolved. Add 700 ml. of boiling water to
the contents of the beaker, stir well, and allow the insoluble
residue to settle for a few minutes. If the supernatant liquid
is violet, due to permanganic acid, add 1 to 2 drops of alcohol,
boil until colourless, and again allow the precipitate to settle.
Decant the liquid through a funne1 containing a large har-
dened filter paper; it is better to allow the latter to rest on a
What man ·filter cone (hardened, No. 51), so that suction may
be applied. Wash the precipitate 4 times by decantation,
using 300 ml. of hot water each time. Transfer the precipi-
tate to the filter, rinsing out the beaker well without making
any attempt to remove the precipitate which adheres to the
• Suitable samples are Ric:h;dale's .. Steel, Nos. I, 2, 3, or 21 .. and" Pyrolu-
site (Mn) Ore, No. 21 .. (Analysed Samples for Students) or the Bureau of
Analysed Samples .. Carbon Steels, . E ' or • R' .. and .. Mang. Ore ' A ' ..
(British Chemical Standards).
374 Quantitative Inorganic Analysis •
glass. Wash the precipitate with hot water until the filtrate
no longer turns starch-potassium iodide paper blue. Place
the filter paper and precipitate in the original beaker, add 50
ml. of freshly standardised O·IN ferrous ammonium sulphate
and 50 ml. of 9N sulphuric acid. Stir the mixture until the
precipitate has entirely dissolved. Titrate the excess of
ferrous salt with standard O·lN potassium dichromate.
Repeat the esti~ation with two other 50 ml. portions of the
solution.
Calculate the percentage of manganese in the sample.
1 Ml. N' K'2Cr:i07 == 0·02747 g. Mn.
OXIDATIONS WITH CERIe SULPHATE
m, 73. General discussion.-As a result of the systematic
investigations, commenced in 1928, of Willard and Young, of
Furman and his co-workers, and of Atanasiu, the use of eerie
sulphate as a volumetric oxidising agent has become firmly
established. Ceric SUlphate is a powerful oxidation agent;
its oxidation potential in 1 to 8N sulphuric acid at 25°C is
1·43 ± 0·1 volts (compare Section I, 46). It can be used only
in acid solution, best in 0'5N or higher concentrations: in
neutral or alkaline solutions, basic ceric or perceric salts
respectively may be formed. The solution has an intense
yellow colour, and in hot solutions which are'not too dilute,
the end point may be detected without an indicator; the
latter procedure, however, nec~ssitates the application of a
blank correction, and it is therefore preferable to add a suit-
able indicator. It should be noted, however, that the self-
colour of ceric solutions is less pronounced than that of
permangana!e solutions, and the titration with an oxalate
cannot be so readily carried out (compare Section m, 76).
The advantages of eerie sulphate as a standard oxidising
agent are:
1. Cerie sulphate solutions are remarkably stable over
prolonged periods. They need not be protected from light,
and may even be boiled for a short time without appre-
ciable change in concentration. The stability of sulphuric
acid solutions coveres the wide range of 10 to 40 ml. of con-
centrated sulphuric acid per litre. It is evident, there-
.. fore, that an acid solution of ceric sulphate surpasses a per-
manganate solution in stability.
2. Ceric sulphate may be employed in the determination
of reducing agents in the presence of a high conceJ?-tration of
hydrochloric acid (contrast potassium permanganate, Section
m,46).
3. Ceric solutions in o·lN .solution are not too highly col-
oured to obstruct vision when rell.dihg the meniscus in bur-
ettes and other volumetric apparatus.
4. It was formerly generally assumed that the reaction
involved the simple valency change:
Ce++++ + E .= Ce+++;
the equivalent weight of a ceric salt is therefore the mole-
cular weight. With permanganate, of course, a number of
375
reduction products are possible according to the experi~
mental conditions.
Recent work of G. F. Smith and Getz (1938) upon the single electrode
potentials of the eerie, cerous system in the presence of varying concentrations
of sulphuric, nitric and perchloric acids has yielded the following results:
Acid Concentra-
,
lion Single Electrode Potential Values (Volts)
N H 2 SO, HN0 3 HC10 4

I 1·44 1·61 1·70
2 1"-44 1·62 1·71
4 1·43 1-61 1·75
6 - - 1,82
8 1·42 1·56 1·87
Since the single electrode 'potential varies considerably with the anion
of the acid used but only slightly with its concentration, these authors con-.
elude that the results support the view that the following equilibria are present
in the three solutions respectively:
[Ce(SO,l.--]
[Ce++++] X [SO, P
No measurements are at present available of th~ instability constants of the

various ions (Section I, 12), but it would appear that the ionic equilibria pre-
sent, say, with cerie sulphate in sulphuric aeid solution could be represented
by:
H,[Ce(50.).J ~ 2H+ + [Ce(SO.l.r - ~ 2H+ + Ce++++ + 350.--,
so that the actual reaction which takes place in the presence of a reducing agent,
particularly for cerie sulphate in sulphuric acid solution whi"h has been most
extensively studied, is indeed :
Ce++++ + f ~ Ce+++,
5. The cerous ion is colourless (compare .colourless mangan~

ous ion from potassium permanganate, and green chromic
ion from potassium diGhromate).
6. Ceric 3ulphate is a very versatile {)xidising agent. It
may be employed in most titrations in which permanganate
has been used and also for other estimations.
7. The indicators which have been used in ceric sulphate
titrations include: diphenylamine, diphenylbenzidine,
so~ium diphenylamine sulphonate, methyl red (with these
four indicators, phosphoric a.cid m~st be present), erio glaucine,
erio green, xylene cyanol FF, ferrom, and N-phenylanthranilic
acid. The last three indicators give the most striking colour
changes, and thus render the determination of the correct
end point a very easy matter: moreover, the addition of
phosphoric acid is unnecessary.
8. Ceric sulphate solutions can be readily standardised
against inter alia arsenious oxide, pure iron, 'and ferrous
solutions, and also indirectly against sodium oxalate (Section
m,76).
For some years it was customary to prepare eerie sulphate
solutions from commercial cerie oxide or cerous oxalate.
This somewhat tedious process was doubtless responsible for
the limited application of eerie sulphate as a volumetric
oxidising agent. Pure eerie sulphate and also the more
soluble eeric ammonium sulphate (composition approxi-
mately Ce(SO,)2,2(NH,)2S0,,2H20) is now marketed at a
very moderate price.* The latter is indeed the cheapest
eerie salt, suitable for volumetric work, at present available
on the market; it suffers, however, from the disadvantage
that it is a hy~rated salt. With the discovery by G. F.
Smith, Sullivan and Frank in 1936 of a method for preparing
pure ammonium hexanitrato-eerate (NH,MCe(N0 3)6], ceric
oxidimetry has reached the stage when it should be available
and used in every laboratory. Two crystallisations of the
inexpensive and r~adily obtainable commercial "ceric am-
monium nitrate" from concentrated nitric acid in the pres-
ence of ammonium nitrate give a product of at least 99·5 per
cent purity; with one crystallisation the purity exceeds 99
per cent, and this substance is sufficiently pure for most
purposes. The above authors propose the substance as a
primary standard, but in view of the ease with which eerie
sulphate solutions may be standardised against arsenious
oxide, etc., the writer prefers to avoid the somewhat lengthy
process of drying the recrystallised eerie ammonium nitrate
in a desiccator over potassium hydroxide in order to remove
the last traces of nitric acid. He converts a known weight
of ceric ammonium nitrate into eerie sulphate by evaporating
with concentrated sulphuric acid, dissolves the resultant ceric
sulphate in dilute sulphuric acid, and then standardises the
solution.t One litre of O'lN ceric SUlphate, prepared from
• Marketed by the G. F. Smith.Chemical Co., Columbus, Ohio, U.S.A. ; the
English agents are F. W. I3erk and Co. Ltd .. 52-54 Leadenhall Street, London,
E.C.3. Also by Johnson, Matthey and Co. Ltd., Hatton Garden, London,
E.C.I.
t Unpublished work in the author's laboratory by S. J. Hart and A. S.
Nickelson (1937) has shown that ceric ammonium nitrate may be dissolved
directly in 2N sulphuric acid, and employed {or many purposes in which ceric,
sulphate is used. However, for some determinations the presence of nitrate
ions may be harmful, and, since the conversion into ceric sulphate is so
simple a process, it is best, as a s-eneral rule, to work with solutions of eerie
sulphate.
recrystallised eerie ammonium nitrate, costs about Is. (or
about 20 cents).
m, 74. Preparation of O'lN eerie sulphate.-Method A.
Dissolve 100 grams of commercial eerie ammonium nitrate in
270 ml. of boiling hot concentrated nitric acid, and then dis-
solve 30 g. of ammonium nitrate in. the solution. Filter the
boiling hot solution through a large Jena sintered glass funnel
(Section n, 12A). and cool the' filtrate in ice or, preferably,
allow it to stand overnight in a covered beaker. Filter off
the separated solid on a similar sintered glass funn~l, and
leave it on the pump for I to 2 hours in order to remove the
excess of nitric acid. The yield is 90 per cenf. Crystallise
the recrystallised eerie ammonium nitrate once more if a
product of very high purity is required; for most purposes, as
already pointed out, a second recrystallisation is unnecessary.
Dry the tecrystallised solid at llO°C for 1·5 hours. The most
convenient method is to spread the solid in thin layers on
clock glasses, and to heat the latter either in a Jena glass air
bath (Fig. 2-13) 'or in the asbestos-lined air bath described in
Section n, 13F. (If perfectly pure ceric ammonium nitrate
is required, the twice recrystallised product must be further
dried il!- a vacuum desiccator (Fig. 2-10, B) over potassium
hydroxide to constant weight.) For routine work, evaporate
55 g. of the dried eerie ammonium nitrate (equivalent and
molecular weight = 548'27) almost to dryness with excess
(48 m1.) of concentrated sulphuric acid either in a Pyrex
evaporating dish or in a shallow (" Bako") beaker. Dis-
solve the resultant eerie sulphate in N sulphuric acid (28 ml.
of concentrated sulphuric acid to 1 litre of water), transfer
quantitatively to a 1 litre volumetric flask, add N sulphuric
acid until near the graduation mark, and then make up to
the mark with distilled water. Shake thoroughly.
Method B. The molecular and also the equivalent weights
of eerie sulphate Ce(SO')2 and eerie ammonium sulphate
Ce(SO,h,2(NH')2S0,,2H:iO are 332·25 and 632·57 respec-
tively. Dissolve 65 to 70 g. of eerie ammonium sulphate in, a
solution prepared by adding 28 ml. of concentrated sulphuric
acid to 500 ml. of water; the mixture is stirred until dissolved.
Transfer to a 1 litre measuring flask and make up to the mark
with distilled water. Shake well.
Alternatively, weigh out 40 g. of anhydrous eerie sulphate
into a 600 ml. beaker, add 56 ml. of 1 : I-sulphuric acid, and
stir, with frequent a~ditions of water and gentle warming,.
until the salt is dissolved. Transfer to a 1 litre volumetric
flask, and,' when cold, dilute to the mark. Shake thoroughly.
Note. These salts need not be of a high grade of purity. The
presence of other rare earths introdUces no complications, since
these metals do not exhibit any variations in valency. If the
salts are impure, it will of course be necessary to increase the quan-
tities given above by 10 to 75 per cent according to the percentage
of cerium present.
'm, 75. Standardisation of eerie sulphate solutions.-Dis-
cussion. The most trustworthy method for standardising
eerie sulphate solutions is with pure arsenious oxide. The
reaction between eerie sulphate and arsenious acid is very slow
at the ordinary temperature; it is necessary to add a trace of
osmium tetroxide or, less satisfactorily, iodine chlqride as a
catalyst. The arsenious oxide is dissolved in sodium
hydroxide solution, the solution acidified with dilute sulphuric
acid, 2 drops of .. osmic acid:' added, followed by the
indicator, and the resultant solution is titrated against the
eerie sulphate solution to the first sharp colour change. As
already pointed out (Section m, 73) the most satisfactory
indicators are: (i) a 0·1 per cent solution of xylene cyanol
FF (0'5 mI.)*; (ii) 0'025M ferroint (lor 2 drops); and (iii)
0·005M N-phenylanthranilic acidt (0'5 mI.). The colour
changes are respectively: (i) yellowish-green to orange; (ii)
orange-red to very pale blue; and (iii) yellowish-green to
purple.
AS 20 3 + 6NaOH = 2NaaAsOa + 3H 20.
2Na aAsO a + 3H2S0 4 = 2HaAs03 + 3NazSO,.
HaAsOs + 2Ce(SO')2 + H 20
= H~As04 + Ce 2 (S04)S + H 2 S0 4 ,
J\nother method is to use pure iron. A known weight, is
dissolved in dilute sulphuric acid, and the resultant ferrous
solution titrated with the eerie sulphate solution; any of the
three indicators, mentioned ~bove, may be employed.
2FcSO, + 2Ce(SO')2 = Fe 2 (SO,)s + Ce 2 (SO,}a.
* The end point is difficult to detect by artificial light with tills indicator.
t This indicator solution may be purchased (see footnote in Section In, 63)
or may be prepared, if desired, as follows. Dissolve 1·485 g. of ortho-phenan-
throline monohydrate C12 H SN Z,H,O (molecular weight, 198; melting point,
90° to 100°C) in 100 ml. of 0·025M ferrous sulphate, prepared by dissolving
0·695 g. of crystallised ferrous sulphate FeSO.,7H,O in 100 mL of water. '
t .Dissolve 1,07 g, of N -phenylanthranilic acid in 20 m!. of 5 per cent sodium
carbonate solution, and dilute to 1 litre.
For work not requiring a very high degree of accuracy, pure
ferrous ammonium sulphate may be used (see remarks in
Section m, 46).
Solutions of eerie sulphate may be standardised against
sodium oxalate in hot dilute sUlphuric ac~d solution. Here
no satisfactory indicator is available at present, because the
known oxidation-reduction indicators are oxidised before
the equivalent point is reached. Contrary to the experience
of other workers, the writer could not obtain satisfactory
results by the use of iodine chloride as catalyst and ferroin as
indicator. The end point may, however, be determined
potentiometrically, or by the addition of eerie sulphate solu-
tion to the hot solution of oxalate to the first appearance of a
pale yellow colour. There is an appreciable blank correction
in .the latter procedure, and the method is not recommended
except for approximate work.
Proeedure A. Standardisation with arsenious -oxide. Weigh
out accurately about 0·25 g': of A.R. arsenious oxide, pre-
viously dried at 105° to llO°C for 1 to 2 hours, and transfer
to a 400 ml. beaker or to a 350 ml. conical flask. Add 20 ml.
of 5 per cent sodium hydroxide sol)ltion, and warm the
mixture" gently until the arsenious oxide has completely
dissolved (I). Cool to room temperature, and add 100 ml.
of water, followed by 25 ml. of 1 : 5 sulphurIC acid. Then
add 3 drops of O·OIM osmium tetroxide solution (0·25 g. of
osmium tetroxide dissolved in 100 ml. of o·lN sulphuric acid)
and 0'5 ml. of N-phenylanthranilic acid indicator (or 0'5 ml.
of xylene cyanol FF, or 1 to 2 drops of ferroin). Titrate with
the O·lN ceric sulphate solution until the first sharp colour
change occurs (see Discussion above). Repeat with two
other samples of approximat~ly equal weight of arsenious
oxide.
-Calculate the normality of the ceric sulphate solution . (for
details, see Section m, 48, Procedure A).
Note. 1. Elementary students may prepare 250 ml. of an approxi-
mately O'lN arsenious oxide solution, and titrate 25 mI. of this
against the eeric solution as described above. Full details for the
preparation of the solution and the method for the calculation of the
result will be found in Section III, 48, Procedure A, Note 1.
Procedure B. Standardisation against pure iron. Weigh
out accurately 0·15 to 0·20 g. of A.R. iron wire, and then
proceed exactly as described in Section m, 48, Procedure C.
Titrate the resultant ferrous solution against the eerie
sulphate solution, using any of the indicators referred to in
the Discussion.
Anotlter method, which incidentally utilises the stannous
chloride process of reduction, is the following. Weigh out-
accurately about 0·20 g. of A.R. iron wire into a 400 m!. beaker
or into a 350 mI. conicalllask. Add 5 mI. of water and 10 m!. of
concentrated hydrochloric acid. Warm the solution until
all the iron has dissolved. Heat the solution to boiling, and
add freshly prepared 0'5N stannous chloride, dropwise and
with constant stirring, from a burette or separating funnel
(Section m, 51A) until the hot solution no longer shows any
yellow colour and a few drops of the reducing agent causes
the colour to change from yellow to very pale green. Cool,
add 5 ml. of cQncentrated hydrochloric acid, dilute the solu-
tion with cold air-free water to 250 m1., and add 10 m1. of a
saturated solution of mercuric chloride. A slight white silky
precipitate pf mercurous chloride should be formed, some-
times after 1 to 2 minutes. (If the precipitate is bulky or
grey in colour, the solution should be rejected, and the esti-
mation repeated.) Add a drop or two of ferroin and titrate
slowly with the' eerie sulphate solution until the sharp colour
change from red to a faint yellow or green colour occurs.
Alternatively, add 5 ml: of syrupy phosphoric acid, 0·5 m!.
of sodium diphenylamine sulphonate, then titrate the mixture
against the ceric sulphate solution (proceed slowly near
the end point) until the solution assumes a distinct purple
tint. Repeat with two other samples of the A.R. iron
wire.
Calculate the normality of the ceric sulphate solution.
Procedure C. Standardisation against ferrous ammonium
sulphate. As pointed out in Section m, 46, this method is
only suitable for work requiring moderate aGcuracy or for
elementary students. Prepare an approximate O'IN solution
of ferrous ammonium s\!lphate (Section m, 48, Procedure D).
Transfer 25 ml. of this solution to a conical flask, add 25 m!.
of N-sulphuric acid, and the appropriate quantity of xylene
cyanol FF, ferroin or N-phenylanthranilic acid (see Discus-
sion above); titrate with the ceric sulphate solution to the
first sharp colour change. Carry out 2 or at most 3 titrations;
these shoufd agree within 0·1 ml.
Alternatively, the ferrous solution may be placed in the
burette, and 25 ml. portions of the eerie solution titrated.
The end points are equally satisfactory, although when. once
the colour of the eerie solution has been bleached, it is advis-
able to add the ferrous solution slowly.
Calculate the normality of the eerie sulphate solution.

CERIC SULPHATE SOLUTIONS
Practically all the determinations described under potas-
sium permanganate and potassium dichromate may be carried
out with eerie sulphate. Use is made of the various indicators
already detailed and also, in some cases where great accuracy
is not required, of the pale yellow colour produced by the
eerie sulphate itself. Only a few determinations will therefore
be considered in some detail.
m. 76. Determination of oxalates (H 2C20 4 in crystallised
oxalic acid).-Discussion. This determination will give the
student practice in the titration of oxalates with eerie sul-
phate ~olution. Two methods may be used. In the first,
the eerie sulphate solution is run into the hot solution of
oxalic acid acidified with dilute sulphuric. acid until the
solution acquires a pale yellow colour, the eerie solution thus
acting as its own indicator.
H 2C2 0" + 2Ce(SO')2 = 2C0 2 + Ce 2 (SO,,)a + H 2S0 4 ,
In the second procedure, a known excess of the eerie solution
is added to the oxalic acid solution, the mixture warmed for
5 minutes at 50°C in order to complete the reaction, and, after
cooling, the excess of eerie sulphate is titrated with ferrous
ammonium sulphate solution, using N-phenylanthranilic
acid or ferroin as indicator.
The latter method provides an excellent means for standar-
dising eerie sulphate solutions. Thus if .roughly equivalent
solutions (say, of about O'IN strength) of eerie SUlphate and
of ferrous ammonium sulphate be titrated, the relative con-
centrations will be found. Now if a known weight of A.R.
sodium oxalate be dissolved in dilute sulphuric acid, excess of
the eerie sulphate solution added, the mixture heated at 50°C
for 5 minutes, and, after cooling, back titrated with the
ferrous ammonium sulphate solution, then the volume of
eerie sulphate solution equivalent to the known weight of
the pure sodium oxalate is obtained. The nqrmality of the
eerie sulphate solution can then be readily calculated.
Procedure. Weigh out accurately about 3·15 g. of A.R.
crystallised oxalic acid (H 2 C2 0",2H 2 0), dissolve it in water,
and make up to 500 ml. in a volumetric flask. Shake well.
Transfer 25 ml. of this solution with a pipette into a conical
flask, add 25 ml. of N sulphuric acid, heat to 70° to 80°C,
and titrate against the O·IN ceric sulphate until the solution
assumes a faint yellow colour. Repeat the titration with
50 ml. of N sulphuric acid, heat to·70° to 80° C, and add the
cerie SUlphate solution until the solution has the same colour
as in the previous titration. Subtract this blank correction
from the volume of O'lN ceric sulphate consumed in the
titration with the oxalic acid solution. Carry out two other
tit rations with the oxalic acid solution; the values should
agree within 0·1 to 0·2 m l . , .
Calculate the percentage of H 2C20" in the oxalic a~id
crystals.
I Ml. N Ce(SO.)2 == 0·04502 g. H2C~aO,.
For the alternative method, prepare an approximately
O·lN soJution of ferrous ammonium sulphate (Section m,
48, Procedure D). Place this in the burette and titrate it
against a mixture of 25 ml. of the O·lN ceric sulphate and 25
ml. of N sulphuric acid, using ferroin (1 drop) or N-phenyl-
anthranilic acid (0'5 ml.) as indicator. Measure out 25 ml. of
the oxalic acid solution into a conical flask, add 50 ml. of
O'IN ceric sulphate and 50 ml. of N sulphuric acid. Warm
the mixture at 50°C for 5 minutes, cool, and titrate the excess
of ceric sulphate with the ferrous solution, employing the
same indicator as in the first titration. Repeat the deter-
mination with two other 25 ml. portions of the oxalic acid
solution; the titrations should agree within 0·1 ml.
Calculate the volume of the 0'1N cerie sulphate that has
reacted with the oxalic acid, and thence the percentage of
H 2C2 0, in the oxalic a~id crystals. -
m, 7'1. Determination of iron in an iron ore.-The proce-
dure is identical with that given in Section m, 68, except that
after the reduction with stannous chloride, addition of excess
of mercuric chloride solution and 200 ml. of 2·5 per cent sul-
phuric acid, I or 2 drops of ferroin or 0·5 ml. of N-phenylan-
thranilic acid is added. The mixture is then titrated with
0'1N ceric sulphate to the first sharp colour change.
Calculate the percentage of iron in the ore.
1 Ml. N Ce(SOJ2 == 0·05584 g. Fe.
ni:, 78. Determination of ferrocyanides (K,[Fe(CN)6J in
crystallised potassium f.errocyanide).-For practice in this
estimation, useA.R. potassiumferrocyanide K,[Fe(CN)6J,3H 20;
commercial potassium ferro cyanide will, however, serve
equally well. Weigh out accurately about 1 g. of the potas-
sium ferrocyanide, and dissolve it in 100 ml. of distilled
water-in a conical flask. Add 25 ml. of 1 : 5 sulphuric acid or
25 ml. of 1 : 1 hydrochloric acid, and I drop of ferroin or 0·5
ml. of N-phenylanthranilic acid. Titrate with standard O·IN
eerie sulphate to the first sharp colour change. Repeat the
determination with two other samples of the potassium ferro-
cyanide.
2K 4 [Fe(CN)6] + 2Ce(S04)2
= 2Ka[Fe(CN)o] + Ce 2(S04)a + K 2S0 4 •
Calculate the percentage of K4[Fe(CN)oJ in the crystals.
I Ml. N Ce(SO,,)z == 0·3684 g. K 4[Fe(CN)6J.
m, 79. Determination of nitrites.-DiscuSS1·On. Satisfac-
tory results are obtained by adding the nitrite solution to
excess of standard O·IN eerie sulphate, and determining the
excess of eerie sulphate with a standard ferrous solution
(compare Section m, 58).
2Ce++++ + N0 2- + H 20 = 2Ce+++ + NO a- + 2H+.
N0 2 in potassium nitrite, or the purity of sodkm nitrite,
preferably of A.R. quality.
Procedure. Weigh out accurately about 1'5 g. of sodium
nitrite and dissolve it in 500 m!. of boiled-out water in a
volumetric flask. Shake thoroughly. Place 50 ml. of stan-
dard O·IN eerie sulphate in a conical flask, and add 10 ml. of
4N sulphuric acid. Transfer 25 ml. of the nitrite solution
to this flask by means of a 'pipette, and keep the tip of the
pipette below the surface of the liquid during the addition.
Allow to stand for 5 minutes, and titrate the excess of eerie
sulphate against standard O'IN ferrous ammonium sulph.ate,
using ferroin or N-phenylanthranilic acid as indicator.
Repeat the titration with two further portions of the nitrite
solution. Standardise the ferrous solution by titrating it
with 25 m!. of the eerie solution in the presence of dilute
sulphuric acid (Section III, 750).
Determine the volume of the standard eerie sulphate solu-
tion which has reacted with the nitrite solution, and therefrom
calculate the purity of the sodium nitrite employed. .
1 Ml. N Ce(SO,,)z == 0·02301 g. N0 2.
rn, 80. Determination of tellurium. Method A. A measured
excess of standard O·IN cerie sulphate is added to the tellu-
rous solution (200 m!.) containing 10 m!. of concentrated
hydrochloric acid and about 0·05 g. of chromium sulphate as
catalyst. * The solution is boiled for 10 minutes, then cooled
and back titrated with standard O'IN ferrous ammonium
sulph~te, using N-phenylanthranilic acid or ferroin as indi-
cator. The tellurium is oxidised from the quadri- to the sexa-
valent stage. Selenium does not interfere.
1 Ml. N Ce(S04)2 == 0·06381 g. Te.
Method B. The solution of tellurous acid or of a tellurite
(200 m!.) is treated with 15 m!. of concentrated hydrochloric
acid and a measured excess of standard O'IN potassium
dichromate, and allowed to stand for 30 minutes. The
residual potassium dichromate is then determined by the
addition of excess of standard O'IN ferrous ammonium sul-
phate and back titration with O'IN potassium dichromate,
using N-phenylanthranilic acid or ferroin as indicator (see
Section m, 66). Selenium does not interfere.
1 Ml. N K 2Cr2 0 7 == 0·06381 g. Te.
Method C. This is carried out similarly to Method A under
Selenium (Section m, 62). The solution should contain 0·15
to 0·20 g. of teHurous oxide or its equivalent.
1 Ml. N KMnO" == 0·06381 g. Te.
rn, 81. Determination of thallium.-An iodine chloride
indicator is prepared by dissolving a small crystal of iodine
in 5 m!. of carbon tetrachloride in a stoppered bottle, adding
50 ml. of concentrated hydrochloric acid and then O'IN ceric
sulphate (Section m, 74) dropwise with frequent shaking
until the colour of the iodine just disappears from the solvent
layer. A measured volume of the thallous soilltion is then
added, and standard O·IN ceric sulphate run in as before until
the carbon tetrachloride again becomes colourless. More
hydrochloric acid should be added, if necessary, to maintain
its concentration about 5N.
The change is :
TI+ - 2 E = Tl+++,
whence 1 ml. N Ce (SO')2 == 0·1022 g. Tl.
• Ceric sulphate alone does not oxidise selenite or tellurite, but it oxidises
chromium to the sexavalent stage and this, in tum, oxidises tellurite (but not
selenite) to the sexavalent condition; any Cr vi at the end is reduced by the
ferrous sulphate. The cerie ion acts as a potential mediator (compare Section
I. 6OA).
o
Note. Ceric solutions are also available for the analysis of
hydrogen peroxide or any per-salts such as persulphates,
" perborates " and percarbonates; also for any ferrous titra-
tion where potassium dichromate and potassium perman-
ganate are applicable, and for the determination of tervalent
antimony and mercurous mercury.
m, 82. Determination of cerium.-Method A. The cerous
salt in the form of sulphate in 100 mL of 1 : 4 sulphuric acid
is treated with 2 g. of ammonium sulphate, 1 g. of A.R.
sodium bismuth ate is added, and the solution heated to boil-
ing. The mixture is cooled somewhat, 50 mL of 2 per cent
sulphuric acid added, filtered through a Gooch, porous
porcelain or sintered glass crucible, and washed with 100 to
150 ml. of 2 per cent sulphuric acid.
2Ce z(SO,)a + 2NaBiO a + 6H 4SO,
= 4Ce(S04)2 + Bi2(SO~)3 + Na 2SO, + 6H20.
Excess of 0'025N ferrous ammonium sulphate is added (as
shown by the change from yellow to colourless, 'and the
consequent complete reduction of the ceric to the cerous salt),.
and the excess of ferrous salt titrated with O· ~N potassium
permanganate to the first appearance of a pink colollr.
1 Ml. 0·025N FeS0 4 == 0'003503 g. Cel
Method B. 100-300 Ml. of the solution, containing 0·1 to
0·3 g. of cerium and 2'5 to 7·5 ml. of concentrated sulphuric
acid, are treated with 1 to 5 g. of A.R. ammonium persulphate
and 10 drops of O'IN silver nitrate solution (catalyst), and
then boiled for 10 minutes. The solution is cooled to room
temperature, and is ready for titration. Two procedures may
be used.
(i) Add 10 to 20 mL of 10 per cent potassium iodide solu-
tion, and titrate the liberated iodine with O'lN or 0·025N
sodium thiosulphate. To avoid the oxidation of the hydriodic
acid by the air, the titration should be performed in an
atmosphere of carbon dioxide.
1 Ml. N Na 2S20 a = 0·1401 g. Ceo
(ii) Titrate the solution with O'IN or 0·025N ferrous
ammonium sulphate, using xylene cyanol FF, N-phenylan-
thranilic acid or ferroin as indicator (Section m, '(5).
1 Ml. N FeS0 4 == 0·1401 .g. Ceo
OXIDATIONS WITH MANGANIC SULPHATE
m, 83. General discussion.-The introduction of aqueous
manganic sulphate as a volumetric oxidising agent by
A. R. J. P. Ubbelohde (1935) originated in an attempt to im-
prove the poor end points which are sometimes obtained with
potassium permangan;lte solutions, due to the slow rate of
oxidation by the permanganate ion and to other causes. It
was found that reactions involving only the removal of
electrons were more rapid and yielded better end points, for
example: .
Fe++ + Mn+++ = Fe+++ + Mn++.
Manganic sulphate solutions gave satisfactory end points in
the determination of ferrous salts in the presence of chloride
ions (contrast potassium permanganate), nitrites, hydrogen
peroxide, oxalates, and vanadous salts: \
m, 84. Preparation and standardisation of manganic sul-
pbate solunons.-Preparation. This is based upon the oxida-
tion of manganous sulphate with potassium permanganate:
8MnS04 + 2KMn0 4 + 8H 2 S04
= 5Mn 2 (S04)a + K 2S04 + 8H 2 0.
Owing to the deep red tint and also the danger of hydrolysis
in dilute solutions, it is best to work with ca. O'07N solutions
prepared in the following manner.
To 50 ml. of a solution of A.R. manganous sulphate (15'l g.
in 1 litre of 6N sulphuric acid) add 3 ml. of concentrated
sulphuric acid with water-cooling; then add 12 ml. of O'5N
potassium permanganate solution, 2 ml. at a time, at intervals
of about 3 minutes. After 8 m!. and after 12 ml. of the
permanganate solution have been added, cautiously introduce
a further 2 ml. of concentrated sulphuric acid. Store the
solution for 4 hours before use. With the above precautions,
reduction proceeds smoothly to manganic sulphate, and
higher valency stages never accumulate ill sufficient con-
centration to give rise to troublesome precipitates of higher'
oxides. If larger quantities of the solution are to be pre-
pared, special precautions to ensure adequate cooling are
necessary.
The solution of manganic sulphate has a deep red colour.
It must be stored in blue bottles and kept in a dark cupboard
when not in use. On exposure to light in a clear bottle, a
387
precipitate gradually appears owing to the "formation and
hydrolysis of quadrivalent manganese salts:
2Mn+++ = Mn++ + Mn++++.
Standardisation. This is effected by running the manganic
sulphate solution into standard ferrous ammonium sulphate
solution (Section m, 48, Procedure D); one drop in excess of
the former gives a pink end point with a yellowish tinge. A
better pink colouration, particularly for use in artificial light,
is obtained on adding 2 drops of syrupy phosphoric acid.
m, 85. Applications of manganic sulphate solutions.-A.
Ferrous salts in the presence of chloride ions. Satisfactory results
are obtained up to a concentration of hydrochloric acid of N.
When the concentration of acid exceeds 0·2N, 0·5 to 1 ml. of
syrupy phosphoric acid is added to improv~ the end point
and the titration is carried out slowly and with vigorous
sqaking.
B. Nitrites. Excellent end points are obtained and the
tit rations are rapid (contrast Section III, 58).
C. Hydrogen peroxide. Manganic sulphate solution reacts
almost instantaneously with hydrogen peroxide, and the end
point is sharp.
D. Oxalates. Titration is effected at 35°C, and the end point
is sharp.
E. Vanadium. Vanadium salts are reduced with liquid
zinc amalgam or with amalgamated zinc to the bivalent stage;
the latter solution at 45° to 50°C may be rapidly and satis-
factorily oxidised by manganic sulphate solution to the quin-
quevalent stage. This procedure is superior to the perman-
ganate method (Section m, 94).
REDUCTIONS WITH TITANOUS SALTS
m, 86. General discussion.-E. Knecht (1903) first sug-

gested the use of titanous chloride for the determination of
iron. This estimation differs from those already described
in that the iron is reduced instead of oxidised, for example:
FeCl 3 + TiCla = FeClg + TiCl4 •
.
The end point of the reaction is judged by the use of an indi-
cator such as methylene blue, which is reduced and decolour-
ised when the titanous solution is present in slight excess, or
of ammonium thiocyanate, which remains red as long as any
ferric salt is present. The latter is generally preferred
because methylene blue is somewhat slow in reaction below
about 35°C; if, however, one drop of 10 per cent sodium
salicylate solution is added, the cha:Qge is instantaneous at
the ordinary temperature.
Titanous chloride is a more powerful reducing agent than
stannous chloride (compare Table XVI). It is oxidised rapidly
in the air and should be kept out of direct sunlight. A. S.
Russell. (1926) has shown that titanous sulphate in 4N
sulphuri<; acid solution is more stable to air than is the
chloride (compare ferrous chloride and fertous sulphate),
and, indeed, its titre against potassium permanganate solution
when kept in a burette exposed to air remains unchanged
for 12 hours. Nevertheless, for accurate work and for
requiring a large number of routine determinations, solutions
of both titanous chloride and sulphate should be stored
in an atmosphere of hydrogen; titrations, particularly
of -the chloride, are best carried out in an atmosphere of
carbon dioxide.
Formerly titanous salts were employed for .numerous quan-
titative estimations*; many of these determinations can now
be made more simply by other methods. The chief present
uses of the reagent are in the estimation of ferric salts, of
organic nitro and nitroso compounds, azo dyes, and various
routine analyses. Only the first-named falls within the
.. For complete details, see E. Knecht and E. Hibbert, New Reduction
Methods of Volumetric Analysis, 1925 (Longmans, Green and Co.), W. M.
Thornton, Titanium, 1927 (Chemical Catalog Co.).
389
390 Q1tantitaf,ive Inorganic Analysis
compass of this volume. The reaction with a nitro compound
may be written:
R.N0 2 + 6Ti+++ + 6H+ = R.NH z + 6Ti++++ + 2H 2 0,
hence N0 2 ;::; 6Ti+++*.
m. 87. Preparation of
titanous solutions.-A. Titanous
chloride solution. Boil 60 ml. of commercial titanous chloride
solution (15 to 20 per cent) for I minute with 100 ml. of con-
centrated hydrochloric \acidt in a small flask, allow to cool,
and make up to about 2 litres with boiled-out water in the
storage bottle (Fig. 2-25). The solution should be thoroughly
mixed by shaking and, when freshly prepared, should fill
the bottle to the neck; otherwise air remains, which must be
completely displaced by hydrogen.
The burette is first completely filled with the solution.
The storage bottle, etc., is connected to a Kipp's apparatus
generating hydrogen. The tap of the burette is opened so
that the latter is emptied and hydrogen is passed through the
system in order to displace the residual air. The tap is now
closed, and the burette filled with the s_olution by.suitably
rotating the tap. It is now ready for standardisation and
for use.
B. Titanous sulphate solution. This is similarly prepared
by boiling 60 ml. of commercial titanous sulphate solution
(15 to 20 per cent) for 2 to 3 minutes with 200 mI. of I : 3
sulphuric acid in a small flask, allowing to cool, and diluting
exactly as detailed under titanous chloride.t If 3 to 4N sul-
phuric acid is used for dilution, the stability of the solution is
considerably increased. Another method, due to A. S. Russell
(1926), is to reduce a solution of titanic sulphate in.4N sulphuric
acid with liquid zinc amalgam (Section m .. 95) for 1 to 2
minutes, and then to decant the resultant titanous sulphate
solution into a storage vessel or flask. This method is par-
ticularly valuable when relatively small volumes of the solu-
tion are reqnired.
* For further details, see J. F. Thorpe and M. A. Whiteley, A Student's
:11anualof Volumetric Analysis, 1925 (Longmans, Green and Co.) and W. M.
W. M. Cumming, I. V. Hopper and T. S. Wheeler, Systematic Organic Chemistry,
1937 (Constable and Co.).
t Commercial titanous chloride solution sometimes contains sulphides, which
are eliminated as hydrogen sulphide on boiling.
t CaU:.n and Henderson (1920) use a more concentrated solution of titanous
sulphate for the titratIOn of nitro and other organic compounds. It is pre-
pared by bOIling 300 mI. of commercial tJtanous sulphate solution (15 to :W
per cent) with 5UO ml. of 1 : 3 sulphuric acid for a few minutes, cooling, and
making up to 1 litre.
In,88. Standardisation of the titanous sulphate (or titanous
chloride).-A. With pure iron. Weigh (lut accurately 1·001
g. of A.R. iron wire (assay va1ue 99·9 per cent; see Section
nI, 46) into a long-necked flask held in an inclined position,
or into a conical flask carrying a short funnel in the mouth,
and add ab_out 100 ml. of 5N sulphuric acid. When all the
iron has passed into solution, transfer quantitatively to a 500
ml. volumetric flask, add a further 200 ml. of 5N sulphuric
acid, and make up to the mark. Shake thoroughly. Pipette
25 ml. of this solution (;:::; 0·0500 g. Fe) into a 250· to 350 ml.
conical flask, fitted with a three-holed rubber stopper carrying
tubes as described in Section m, 116B), and carefully oxidise
with potassium permanganate solution of about 0'02N
strength until a faint pink tinge is obtained. Boil vigorously
to expel dissolved oxygen whilst a slow current of carbon
dioxide is passed through the fla~k. Cool the solution, add 10.
to 20 ml. of 10 per cent ammonium or potassium thiocyanate
solution and run in the titanous sulphate (or chloride)
solution until the red colour, due to " ferric thiocyanate," has
disappeared. The stream of carbon dioxide is maintained
throughout the whole of the operation. The reduction of the
last few mg. of the ferric salt is somewhat slow and it is
therefore best to add the titanous solution dropwise with
intervals of several seconds between the drops when the end
point has almost been reached. The use of a large excess of
thiocyanate renders the end point much sharper. It is
immaterial whether the iron is present in sulphuric or hydro-
chloric acid solution, but the presence of some mineral acid is
essential, as otherwise the indicator is not sensitive. Repeat
the titration with two other 25 ml. portions of the iron solution.
Calculate the weight of iron equivalent to 1 ml. of the titan-~
ous solution.
B. With ferrous ammonium sulphate. Weigh out accur-
ately 3·511 g. of A.R. ferrous ammonium SUlphate (Section
m, 46) dissolve it in water, add about 150 ml. of :iN sulphuric
acid, and make' up the solution to 250 ml. in a standard flask.
Shake thoroughly. Pipette 25 ml. of this solu~ion (== 0·0500
g. of Fe) into a 250 to 350 ml. conical flask and proceed as
described under A. This method is less accurate .than that
employing pure iron (for reasons, see Sections m, 46 and
m, 48, Procedure D).
A.R. ferric axp.monium sulphate (iron alum) may also be
used for the standardisation.
392 Qttantitative Inorganic 4nalysis
m, 89. Determination of iron in an ore.-Weigh out
accurately about 2 g. of the iron ore, * 'dissolve it jn 100 m!. of
1 : 1 hydrochloric acid, and make up the solution to 250 ml.
in a volumetric flask as described in Section m, 53. Pipette
25 ml. of this solution into a 250 m!. conical flask, and pre-
cipitate. the iron by the addition of ammonium hydroxide
solution. Oxidise the ferrous hydroxide by adding excess of
pure hydrogen peroxide solution. Boil for 10 minutes to
decompose the excess of hydrogen peroxide. Cool, add
concentrated hydrochloric acid until the acid content of the
solution is about 10 per cent by volume, then 10 m!. of 10 per
cent ammonium thiocyanate solution. Titrate against tne
standardised titanous sulphate (or chloride) solution in an
atmosphere of carbon dioxide until the red colour just dis-
appears (Section m, 88A). Repeat with two other 25 m!.
portions of the solution.
From the results, calculate the percentage of total iron in
the sample of ore used.
*Ridsdale's" Iron Ore, No. 17 .. (one of the Analysed Samples for Students)
or the Bureau of Analysed Samples" Iron Ore' A'" (one of the British
Chemical Standards) may be used.
REDUCTIONS WITH AMALGAMATED ZINC AND WITIJ
LIQUID AMALGAMS
m,90. General discussion.-The use of amalgamated zinc
as a reducing agent has been fully dealt with in Section
m, 5IB: full details are given in that Section of the manipu-
lation of the very convenient Jones' reductor. Experimental
details, based upon the use of the Jones' reductor, for the
determination of uranium, titanium, molybdenum, and
vanadium are given below.
Nakazono (1921) introduced liquid zinc amalgam as a
reducing agent in volumetric analysis, and subsequent
Japanese workers, particularly Someya (19.24-1927), have
added liquid cadmium, bismuth and lead amalgams. These
amalgams have different reduction potentials (see Table XV) ;
the order is zinc, cadmium, lead, bismuth, the first-named
being the most powerful reductant. The final reduction
products obtained wit!t these amalgams for a few elements are
collected in the, Table.
Liquid
Amalgam Iron I Titanium I Molyb-
denum
vana-I
I dium
u,ran-I Tung-
,um sten
Zinc Fe++ Ti+++ Mo+++ y++ J U+++~ -
_Cadmium
Lead
Fe++
Fe++
Ti+++
Ti+++
Mo+++
Mo+++
y++
y++
U++++
U++++
-
W+++
Bismuth Fe++ Ti+++ Mo+++ or vo++ U+;H+ W++++
Mo+++++·
Zinc amalgam also reduces chromium to Cr++.

In v.iew of the poisonous character of mercury vapour-
the effect is a cumulative one-the author prefers to recom-
mend the use of the Jones reduct or for most reductions in
routine work. However, in determinations, such as those of
titanium (with zinc amalgam) and of tungsten (with lead.
amalgam), liquid amalgams offer certain advantages, and
their use for these determinations is accordingly described
below (Sections m, 95, 96). These Sections incidentally
contain the methods. of, preparation of liquid zinc and lead
amalgams; the amalgams of cadmium and bismuth are
prepared similarly to that of zinc, except that for bismuth
hydrochloric acid replaces sulphuric acid. Zinc amalgam is
useful in the preparation of titanous, vanadous and chro:qlOus
salt solutions .
• The exact product depends upon the pH of the solution.
0* . 393
A special apparatus is described tor the determination of
titanium (Section m, 95). For many purposes (e.g., the
reduction of ferric salts), however, the reduction may be
carried out in a 250 ml. reagent bottle fitted with a well-
ground stopper. For the reduction of 50 mI. of a solution of
approximately deci-normal concentration, which should be 2
to 4N with respect to sulphuric acid, 30 to 50 mI. of amalgam
may be used; the reaction is complete after shaking for a
few minutes. The contents of the bottle are then transferred
to the titration flask or other vessel, the bottle rinsed 2 or 3
times with small volumes of boiled-out water, the washings
being of course added to the flask. The same amalgam can
be used many times before it becomes exhausted.
ro, 91. Determination of uranium.-The uranium solution
(100 mI., about 0·25 g. U) should contain 5 per cent by volume
of sulphuric acid, no compounds that are reduced in the Jones
reductor, and be free from nitrie acid. Add enough O·IN
potassium permanganate to produce a permanent pink tint,
cool the solution to 20° to 25°C, and pass it through a Jones
redudor (Section III, 51 B) at a rate of 50 to 100 ml. per
minute. Part of the uranium is reduced below the quadri-
valent condition. Bubble a stream of air through the solution
for 5 to 6 minutes; the dirty dark green colour -changes to
the bright apple-green characteristic of uranous salts. Titrate
the solution with standard O'IN potassium permanganate.
Carry out a blank determination with the reagents in the usual
way, and subtract the blank correction from the volume of
potassium permanganate required in the estimation.
I Ml. N KMn0 4 = 0·1190 g. U.
The cold sulphuric acid solution of quadrivalent uranium
is stable to air for several hours, and even a rapid stream of air
effects no change in 30 minutes.
m, 92. Determination of titanium.-The titanium solution
should -be free from all metals that are reduced in a Jones
reductor and also from all substances which are likely to
react subsequently with potassium permanganate. 150 Ml.
of the solution, containing sulphuric acid and 0·12 to 0·15 g.
of titanium, are treated with enough O'IN potassium per-
manganate to produce a permanent pink tint, and then
passed through a Jones reductor (Section m, 51B); the
reduced solution is collected in a five-fold excess of an acidi-
fied solution of A.R. ferric alum-IOO ml. of a 10 per cent
Volumetrt"c Analysis 395
solution suffices for most purposes. The ferrous iron thus
produced by reaction with the tervalent titanium is titrated
with standard O·IN potassium permanganate. Subtract the
blank correction determined in the usual manner.
. 1 Ml. N KMn0 4 == 0'04790 g. Ti.
III, 93. Determination of molybdenum.-The molyb-
denum solution should be free from all metals that are reduced
in a Jones reductor, and from all substances which are likely
to react subsequently with the potassium permanganate.
The cold solution, containing 0·08 to 0·10 g. of molybdenum
and 3 to 10 per cent by volume of sulphuric acid, is treated
with enough deci-normal potassium permanganate to produce
a pink colour, and then passed through a Jones teductor ;
the reduced solution is collected in a five-fold excess of
acidified A.R. iron alum solutio~.(say, 30 ml. of a 10 per cent
solution). Add a few ml. of syrupy phosphoric acid to the
contents of the receiver and titrate the ferrous sulphate pro-
duced with standard O·IN potassium permanganate. Subtract
the volume of permanganate required in a blank determination
in which the reagents have been carried through all the steps
of the process.
1 M1. N KMn0 4 == 0·0320 g. Mo.
Note. . Owing to the instability of tervalent molybdenum com-
pounds, exposure to air should be reduced to a minimum, and the
lower tube of the reductor should pass under the surface of the iron
solution.
In, 94. Determination of vanadium.-MetHod A. The
experimental details are similar to those described under
Titanium (Section III, 92). The vanadium solution, contain-
ing sulphuric acid, is treated with enough O'lN po.tassium
permanganate to produce a: pink colour and is passed through
a Jones reduct or, whereupon it is reduced to the bivalent
stage; the reduced solution is caught in excess of an acidified
solution of A.R. iron alum, and the ferrous iron thus formed is
titrg_ted with standard O'lN potassium permanganate.
Actually, the ferric salt oxidises the van ado us salt only to
the quadrivalent condition, which is stable in air, and the,
final oxidation to th.e quinquevalent state is completed
somewhat slowly by the potassium permanganate. The
titration {s somewhat slow towards the end, but the final
point is quite definite; by heating the solution to 70° to 80°C
toward's the end of the titration, the oxidation is rapid.
1 M1. N ,KMn0 4 ~ 0·01698 g. V.
396 Quantitative Inorganic A'nalysis
Method B. The vanadium solution should contain 2 per
cent by volume of sulphuric acid, and be free from a.rsenic,
antimony, iron, chromium and other substances that are
oxidised by potassium permanganate after reduction with
sulphur dioxide. Heat the solution to boiling,. add a strong
sol}1tion of potassium permanganate until the solution is
pink, fit the flask with wash bottle tubes, and pass sulphur
dioxide through the solution for 5 to 10 minutes; the solution
acquires a pure blue colour (vanadyl sulphate) and is quite
stable in the presence of sulphuric acid. Next pass a rapid
stream of carbon dioxide' through the boiling solution until
it is free from sulphur dioxide, as tested by bubbling the
issuing gas through an acidified and very dilute solution of
potassium permanganate. Cool the solution to 60° to 80°C,
and titrate with standard O'lN potassium permanganate .
. 2H 3V0 4 + H 2S0 3 + H 2S0 4 = 2(VO)S04 + 5H zO;
5V 20 2++++ + 2MnO,- + 22H 20
= lOH sV0 4 + 2Mn++ + 14H+.
J Ml. N KMnO, == 0·05095 g. V.
Note. Since sulphur dioxide does not reduce molybdic acid, this
procedure can be employed in the presence of molybdenum. Both
molybdic and vanadic acids are reduced in the Jones reductor, hence
the former may be determined by reducing both wi~ a,malgamated
zinc and making allowance for the vanadium present.
m, 95. Determinaijon of titanium (zinc amalgam method)
-This method is based upon the use of liquid zinc amalgam
as the reducing agent, whereby the titanium is quantitatiVely
reduced to the tervalent state; the latter
is titrated against standard ferric ammo-
nium sulphate solution with potassium
250 mi. thiocyanate solution as indicator.
The apparatus employed for the titration
is shown in Fig. 3-7. A 250 m!. globular
separating funnel is closed with a rubber
stopper A carrying a piece of 0·25 inch
tubing B, which is closed by a small cork C.
(Alternatively, 'the funnel may be closed
with a three-holed rubber stopper carrying
inlet and outlet tubes for carbon dioxide
and a small stopper in the third hole.) The
lower end of the funnel is connected by
means of thick-walled (" pressure ") tUbing
FIG. 3-7. E to a 25 ml. flask D (this may consist of
a 25 ml. volumetric flask, etc.); F is a pinch cock.
A sample of the substance, equivalent to 0·1 to 0·2 g. of
Ti0 2 (1), is digested with 20 ml. of concentrated sulphuric
acid and 15 g. of powdered A.R. ammonium sulphate until it
is entirely dissolved. Bulb D and the rubber tubing, etc.,
to the stop-cock G are filled: with cooled, freshly boiled dis-
tilled water containing I per cent of sulphuric acid (" acid
distilled water "), and F and G are closed. Fifteen ml. of
liquid zinc amalgam are placed in the funnel, and the cooled
sample is transferred to the funnel using about 75 m!. of
"acid distilled water." Several grams of A.R. sodium
bicarbonate (preferably in th,e form of two tablets) are added,
and stopper A is inserted with cork C removed. When the
gas evolution has ceased, a further few grams of A.R. sodium
bicarbonate are added through B. When effervescence
ceases, cork C is immediately replaced and the whole appara-
tus vigorously shaken for 5 minutes. The funnel is then
supported in a stand, G and F are opened, and the amalgam
is allowed to flow into D: this is best accomplished by alter-
natively squeezing and relea!>ing the tube E with the fingers.
When the last particles of amalgam have dropped into the
flask, G and F are closed, and bulb D is disconnected for
convenience whilst titr~ting. Stopper C is removed, and 5 ml.
of saturated potassium thiocyanate solution added. through
B by means of a pipette. Stopper A is removed and both
tube and stopper are washed into the funnel with "acid
distilled water." The solution is then titrated in the funnel
with the ferric ammonium sulphate solution (2). It is impor-
tant to add the latter solution very rapidly until the first
appearance of a wine-red colour. Stop-cock G is then opened,
and tube E squeezed several times to force its liquid into the
funnel; this will cause the red colour to disappear. The
titration is then completed by adding the ferric solution drop
by drop to the end point.
The ferric ammonium sulphate solution is approximately 0·07N
and is prepared by dissolving ca. 30 g. of the A.R. salt in a mixture
of 300 ml. of water and 10 m!. of concentrated sulphuric acid;
potassium permanganate solution is added dropwise as long as the
pink colour d,isappears (if at all), and the solution is then diluted to
1 litre. The solution is standardised, after reduction with zinc
amalgam in the above apparatus, by titration against standrrd
O'IN potassium permanganate.
The dnc amalgam is prepared by washing 15 g. of pure, fine-mesh
zinc (e.g., A.R.) with dilute sulphuric acid, and then heating for 1
398 Quantitative Inorganic Analysis,
hour on the water bath with 300 g. of mercury plus 5 m!. of 1 : 4
sulphuric acid. (Caution. Mercury vapour is highly poisonous;
the operation must therefore be performed in a fume cupboard
provided with a good draught.) The whole is allowed to cool, the
amalgam washed several times with dilute sulphuric acid, and the
liquid portion separated from the solid by means of a separating
funnel. The solid is reserved f9r another preparation of the
amalgam. The liquid amalgam is preserved under dilute sulphuric
acid; reaction with the latter is very slow, and the same sample of
amalgam may be employed for several reductions.
Notes. 1. This method is of particular value for the determination of titanium
in pigments and similar products. Calcium sulphate does not interfete with
the subsequent reduction because its precipitation is prevented by the high
concentration of sulphuric acjd. Barium sulphate should be removed by
filtration l>efore the reduction of the titanium. If two determinations are
carried out, and in one case the rectuced solution be titrated with the ferric
solution and potassium thiocyanate, and in the other case by standard O·lN
potassium permanganate (this gives 'Ii + Fe), the method may be employed
for the determination of iron and titanium in ad"Uixtnre.
2. For the most accur:ate results it is best to insert the three-holed rubber
stopper referred to above, and to conduct the titration in an atmosphere of
carbon dioxide.
3. It is advisable to run 'a blank to determine how much ferric solution is
required to give an end point in a volume of 3 per cent sulphuric acid equal in
volume to that of the final solution and containing 5 m!. of saturated potas-
sium thiocyanate solution.
ITI, 96. Determination of tungsten.-This element is not
reduced to any definite valency stage in the Jones reductor.
Lead amalgam reduces tungsten as tungstate to the tervalent
condition (W 20 3 ) in strong hydrochloric acid at 50°C in an
atmosphere of carbon dioxide (K. Someya, 1925). This is
titrated with standard O'lN potassium permanganate in the
presence of concentrated manganous sulphate ,solution; it
is thereby oxidised to the WO s stage, when 3 equivalents of
KMnO, are equivalent to I mol of W0 3 •
The lead amalgam is prepared by washing pure lead with con-
centrated hydrochloric acid (to remove the film of oxide), and then
heating with pure mercury in a casserole until a homogeneous liquid
is produced. This process should be carried out in a fume cupboard
provided with a very eflj,cient draught. The amalgam is allowed to
cool, washed with cold distilled water, and the solid removed in a
separating funnel by appropriately manipulating the tap.
m, fYI. RE:QUCTIONS WITH CHROMOUS SALTS.
. Chromous solutions ar~ more powerful reducing agents
than titanous salts, and their application to volumetric analy-
l)is has recently been widely studied, pal;.ticularly by Zintl
and his c6-workers (1927 ~). The chromous solution is best
prepared by the foJlowing series of changes :
\10___-" Cre I,
HCl Zn Na.C.H30. _____
K.Cr.O. --+ ercl. --+ CrCI, -* Cr(C.HaO.),.
(solid) ::--
Ras o , ~ CrSO,
Chromous solutions are readily oxidised by air, hence pre-
cautions must be taken in their preparation, storage and use
to exclude oxygen. A suitable apparatus for storing these'
solution is shown in Fig. 2-25 or 2-26. End points are usually
determined by -potentiometric titration. The solutions are
standardised against pure copper sulphate or pure potassium
dichromate.
The r.eagent may be employed inter alia for the determina-
tion of copper, mercury, silver, gold, stannic tin, bismuth,
iron, molybdenum and titanium, and also for the following
mixtures by a single titration: copper-mercury, copper-
silver, copper-iron, copper-tin, gold-copper-silver and gold-
copper-mercury*'
... For further details, the reader is referred to W. Bottger-R. E. Oesper,
.Newer Methods of Volumetric Chemical Analysis, 1938, pp. 131-l52 (D. Van
Nostrand Company; Chapman and Hall).
399
m, 98. REDUCTIONS WITH VANADOUS SALTS
Vanadous salts are more powerful reducing agents than
titamous salts but less powerful' than the corresponding
chromous salts. Their introduction to volumetric. analysis
proper is due to A. S. Russell (1926), who prepared a solution
of vanadous sulphate by reducing aqueous vanadic sulphate
with liquid zinc amalgam and showed that the solution'in 4N
sulphuric acid is stable for about 1 hour in an open burette.
However, precautions should be taken to exclude oxygen in
all work with vanadous salts.
K. Maass (1934) prepared the reagent by the electrolytic
reduction of ammonium vanadate in the presence of dilute
hydrochloric or sulphuric acid, and applied it to the poten-
~iometric titration of ierric iron, copper, and silver, and to,
mixtures of copper-silver, copper-iron, iron-silver, and iron-
vanadium.
P. C. Banerjee (1935) employed a solution of vanadous
ammonium SUlphate, prepared by the electrolytic reduction
of aqueous vanadic sulphate in the presence of ammonium
sulphate, and standardised it against a ferric solution in the
presence of thiocyanate as indicator. This solution is con-
sidered more stable than a van ado us "sulphate sdlution, and
was employed for the determination of ferric iron, copper,
and chromates.
400
'OXIDATION AND REDUCTION P~OCESSES INVOLVING
IODINE-IODIMETRY AND IODOMETRY
m, 99. General discussion.-Iodimetry covers titrations
with a standard· solution of iodine. Iodometry deals with the
titration of iodine liberated in chemical reactions. The
normal oxidation potential of the reversible system:
12 + 2e ~ 2r,
is 0'535 volt. This shows that iodine is a mucn weaker
oxidising agent than potassium permanganate, potassium
dichromate and eeric sulphate (compare Table XVI). Strong
reducing agents (substances witn a much lower oxidation
potential), such as stannous chloride, sulphurous acid,
hy<)rogen sulphide, and sodium thiosulphate, react com-
pletely ·and rapidly with iodine even in acid solution. With
somewhat weaker reducing agents, e.g., trivalent arsenic,
trivalent antimony and ferrocyanide, complete reaction
occurs only when the solutioa is kept neutral or very faintly
acid; under these conditions the oxidation potential of the
reducing agent is a minimum or its reducing power is a maxi-
mum. The equations for some of the reactions are given
below:
Sn++ + 12 = Sn1-+++ + 2r ;
S03-- + 12 + H 0 = S04- - + 2H+ + 2r;
2
H 2S + 12 = S + 2H+ + 2r;
2S 2 0 3- - + 12 = S400- - + 2r;
HaAsOa + 12 + -H 0 ~ HaAsO" + 2H+ + 2r.
2
If a strong oxidising agent is treated in neutral or (more

usually) acid solution with a large excess of iodide ion, the
latter reacts as a reducing agent and the oXIdant will be
quantitatively reduced. In such cases, an equivalent amount
of iodine is liberated, and is then titrated with a standard
solution of a reducing agent, which is usually sodium thiosul-
phate. Examples of such titrations are illustrated by the
following equa~ions :
2Mnl>4- +l6H+ + lor = 2Mn++ + ~I2 + 8H2 0;
Cr20,- - +14H+ + 6r - 2Cr+++ + 31 2 + 7H 20;
H 20 2 + 2H+ + 21- - 2H 2 0 + 12 ;
BrO a- + fiH+ + fir Br- + 31 2 + 3H 20;
IOs- + 6H+ + 5r 31 2 + 3H 20;
401
CI0 3- + 6H+ + 6r - Cl- + 31 2 +3HP;
ClO- + 2H+ + 2r - Cl- + 12 + H 20 ;
10,- + sH+ + 7r - 412 + 4H2 0;
2HN0 2 + 2H+ + 2r - 2NO + 12 +2H 2 0;
2Ce++++ + 21- - 2Ce+++ + 12 ;
2Cu++ + 41- - CU z12 + 12 ~
Cl 2 + 2r = 2Cl- + 12 ;
Br z +- 2r = 2Br- +" 12,
The normal oxidation potential of the iodine-iodide system
is independent of the pH of the solution so long as the latter
is less than about S; at lower acidities iodine reacts with
hydroxyl ions to form iodide and the extremely unstable
hypoiodite, the latter being transformed rapidly into iodate
and iodide by self oxidation and reduction:
12 + 20H- = r + 10- + H 20;
310- = 2r + IO a-.
It has been shown in Section I, 56 that the oxidation poten-
tials of certain substances increase considerably with increas-
ing hydrogen ion concentration of the solution. This is the
case with systems containing permanganate, dichromate,
arsenate, antimonate, bromate, etc., i.e., with anions which
contain oxygen and therefore require hydrogen for complete
reduction. Many weak oxidising anions are completely
reduced by iodide ions if their oxidation potentials are raised
considerably by the presence in solution of a large amount of
acid.
By suitable control of the pH of the solution, it is sometimes
possible to titrate the reduced form of a substance with
iodine, and the oxidised form after the addition of iodide
with sodium thiosulphate. Thus with the arsenite-arsenate
system:
HaAs03 + 12 + H 2 0 ? HaAsO, + 2H+ + 2r,
the reaction is completely reversible. At pH values between
4 and 9, arsenite can be titrated with iodine solution. In
strongly acid solutions, however, arsenate is reduced to
arsenite and iodine is liberated. Upon titration with sodium
thiosulphate solution, the iodine is removed and the reaction
proceeds from right to left. - •
With certain weak oxidising agents, complete reduction
with the iodide ion may be difficult; the reaction may be
made suitable for quantitative work by one of the folfowing
methods;
(i) Increase of the iodide ion concentration, as for the
reaction:
2Fe+++ + 21- ~ 2Fe++ + 12 ,
(ii) Increase of the hydrogen ion concentration if this
increa5es the oxidation potential of the system.
(iii) The iodine may be removed from the sphere of reaction
either by boiling and then collecting in an appropriate solvent,
or it may be extracted with an immiscible solvent, such as
chloroform, carbon tetrachloride or carbon disulphide.
(iv) If the reacti0n is a slow one in an acid medium, it is
advisable to replace the air in the titration vessel by carbon
dioxide (this is most simply effected by the addition of 2 or 3
half-gram portions of pure sodium bicarbonate). This is
because the velocity of the reaction:
4r- + 4:H+ + O2 = 212 + 2H zO,
which is extremely slow in a neutral medium, increases with
increasing PH' of the solution; it is greatly accelerated by
sunlight and by various catalysts as, for example, traces of
copper.
In a number of cases, as in the analysis of the higher oxides
of manganese and of lead, it:is necessary to boil the oxidising
agent with hydrochloric acid 'or hydrobromic acid, and to pass
the liberated chlorine or bromine into excess of potassium
iodide solution. Direct boiling of the sample with an acid
iodide solution is not advisable, since the latter is a much
stronger reducing agent than chloride or bromide and will
react with ferric iron, which is a common impurity in these
oxides.
MnO z + 4H+ + 4Cl- = Mn++ + C1 2 + 2Cl- + 2H zO ;
C1 2 + 2r- = 2Cl- + 12 ;
2Fe+++ + 2r ~ 2Fe++ + 12,
m, 100'. Detection of the end point.-A ~olution of iodine
in aqueous iodide has an intense yellow to brown colour.
One drop of O'lN iodine solution imparts a perceptible pale
yellow colour to 100 ml. of water, so that in otherwise colour-
less solutions iodine can serve as its own indicator. The test
is made much more sensitive by the use of 9- solution of starch
as indicator. Starch reacts with iodine in the presence of
iodide to form an intensely bJue coloured adsorption complex
containing both iodine and iodide, which is visible at very
low iodine concentrations. Thus 100 ml. of water containing
0·05 to 0·1 g. of potassium iodide and a few ml. of 0·.5 to 1
per cent starch solution becomes distinctly blue upon the
addition of 1 drop of O·lN iodine. The sensitivity of- the
starch reaction' corresponds to an iodine concentration of 1
to 2 X 1O- 5N at about 20° C in the presence of a small amount
of iodide (> O·OO-IN). The reaction is less sensitive in the
absence of iodide and also decreases in sensitivity with increas-
ing temperatu-re of the solution; thus at 50°C it is about 10
times less sensitive than at 25°C.
In recent years, carbon tetrachloride has been used in
certain reactions instead of starch solution. One litre of
water at 25°C will dissolve 0·335 g. of iodine, but the same
volume of carbon tetrachloride will dissolve about 28·5 g.
Iodine is therefore about 85 times as soluble in carbon tetra-
chloride as it is in water (compare distribution coefficient in
Section I, 71), and the carbon tetrachloride solution is highly
coloured. When a little carbon tetrachloride is added to an
aqueous solution containing iodine and the solution well
shaken. the great part of the iodine will dissolve in the
carbon tetrachloride; the latter will fall to the bottom since
it is immiscible with water, and the colour of the organic
layer will be much deeper than that of the original aqueous
solution. The reddish-violet colour of iodine in carbon
tet'rachloride is visible in very low concentrations of iodine;
thus on shaking 5 ml. of carbon tetrachloride with 50 ml. of
2 X 10-5 N iodine, a distinct violet colouration is produced in
the organic layer. This enables us to carry out many iodo-
metric determinations with comparative ease. The reaction
takes place in 250 ml. glass-stoppered bottles or flasks with
accurately ground stoppers. After adding the excess of
potassium iodide solution and 5 to 10 ml. of carbon tetra-
chloride to the reaction mixture, the titration with sodium
thiosulphate is commenced. At first the presence of iodine in
the aqueous solution will be apparent, and gentle rotation of
the liquid causes. sufficient mixing. Towards the end of the
titra tion the bottle or flask is stoppered and shaken after
each addition of sodium thiosulphate solution; the end
point is reached when the car~on tetrachloride just becomes
colourless. Equally satisfactory results can be obtained
with chloroform.
The end point can also be determined potentiometrically
(Section 1,51).
In iodimetry and iodometry the following solutions inter
alia are used: iodine, sodium thiosulphate, sodium arsenite,
potassium iodate, potassium bromate, and chloramine-T. All
of these will be discussed in the ensuing pages. Mention must,
however, now be made of the use of a standard solution containing
potassium iodate and potassium iodide. The latter is quite
stable and yields iodine when treated with acid:
KIO a + 5KI + 6HCI = 6KCl + 31 2 + 3H 2 0;
or 10 s- + 5r- + 6H+ = 31 2 + 3H 20.
The standard solution i~ prepared by dissolving a weighed
amount of pure potassium iodate in a solution of potassium
iodide, and diluting to a definite volume. This solution has
two important uses. The first as a source of iodine in titra-
tions; it must be added to a solution containing strong 'acid,
but it cannot be employed in a medium which is' neutral or
possesses a low acidity. The second is in the determination
of the acid content of solutions. If excess of the neutral
solution of iodate and iodide .i&added to a dilute solution of a
strong acid, the hydrogen ions of the latter yield an equivalent
amount of iodine, which can be titrated with standard sodium
thiosulphate solution. This method is particularly useful
in the titration of very dilute solutions of strong acids, be-
cause the end point is readily determined. It is not very
satisfactory for weak acids owing to the slowness of the reac-
tion in solutions ,of low acidity; results may, however, be
obtained by adding an excess of sodium thiosulphate solution
after the addition of the iodate-iodide mixture, and then
back-titrating the excess of thiosulphate with standard
iodine solution,
m, 101. Preparation of O'lN sodium thiosulphate.-Dis-
cussion: Sodium thiosulphate Na 2S20 a,5H 20 is readily
obtainable in a state of high purity, but there is always some
uncertainty as to the exact water content because of the'
efflorescent nature of the salt and for other reasons. The
substance is therefore unsuitable as a primary standard.'"
It is a reducing agent by virtue of the reaction:
2S 20 S- - -+ S406 +2 € ;
the equivalent weight of sodium thiosulphate pentahydrate is

its molecular weight, or 248·20t. An approximately O'IN
... A.R. sodium thiosulphate has a purity of 99 to 101 per cent.
t'The same result is deduced from the equations:
I. + 2. = 21-,
and 2Na.S.O. + I. = Na.S.O. + 2NaL
Alternatively, sodium thiosulphate may be regarded as interacting with weak
oxidising agents in accordance with the equation:
2Na.S.O. + 0 + H.O = Na.S.O. + 2NaOH.
l'his leads to the same result, flie.,.the equivalent weight is 1 mol.
solution is prepared by dissolving about 25 g. of A.R. crystal-
lised sodium thiosulphate in 1 litre of water in a volumetric
flask. The solution is standardised by any of the methods
described below.
Before dealing with these, it is necessary to refer briefly
to the stability of thiosulphate solutions. Solutions prepared
with conductivity (equilibrium) water are perfectly stable.
However, ordinary distilled water usually contains an excess
of carbon dioxide; this may cause a sl~w decomposition to
take place with the formation of sulphur: -
SZ03- - + H+ -+ HS0 3- + S.
Moreover, decomposition may also be caused by bacterial
action (thiobacillus thioparus), particularly if the solution has
been standing for some time. For these reasons, it is recom-
mended:
(i) to prepare fhe solution with recently-boiled distilled
water;
(ii) t.o add 3·8 g. of borax or 0·1 g. of anhydrous sodium
carbonate per litre; this serveS the double purpose of pre-
venting the above decomposition and helping to keep the
solution sterile;
(iiiL exposure to light should be avoided, as this hastens
decomposition. !
The standardisation of thiosulphate solutions may he
effected with potassiuIl;t iodate, potassium bromate, potassium
permanganate, potassium dichromate, copper, or iodine.
Owing to the volatility of iodine and the difficulty of prepara-
tion of perfectly pure iodine, this method is not a suitable one
for beginners. If, however, a standard solution of iodine
(see Sections m, 103, 104) is available, this may be used for
the standardisation of thiosulphate solutions.
Preparation and use of starch solution. Make a paste of
1 g. of soluble starch with a little water, and pour the paste,
with constant stirring, into 100 ml. of boiling water, and boil
for 1 minute. Allow·the solution to cool, and add 3 g. of
potassium iodide. This solution may be preserved for a
long period if kept under a layer of toluene in a stoppered /
bottle.
A starch solution which keeps almost indefinitely may be,
prepared by dissolving 1 g. of soluble starch and 0-5 g. of pure
salicylic acid in 200 m!. of water. It is best, as a general rule,
to confine the use of this indicator to elementary work owing
to the possibility of the preservative reacting with one of the
Volumetric. Analysis 407
reacting substances. If, however, it has been ascertained
that the salicylic acid causes no interference in the particular
reaction under study, then the starch-salicylic acid indicator
may, of course, be used*.
If soluble starch is not available, grind 1 g. of ordinary
starch with a little cold water, pour the paste slowly into 100
ml. of boiling water, and boil the solution for 2 minutes. Cool
by placing into cold -water, and filter the cold solution. This
solution cannot safely be kept for more than about a day.
Only freshly prepared. or properly preserved starch solution
should be used. Two ml. of a 1 per cent solution per 100 IIlI.
of the solution to be titrated is a satisfactory amount; the
same volume of starch solution should always be added in a
titration. In the titration Of iodine, starch must not be
added until just before the end point is reached. Apart from
the fact that the fading of the iodine colour is a good indication
of the approach of the end point, if the starch solution is
added when the iodine concentration is high, some iodine may
remain adsorbed even at the end point. The indicator blank
is negligibly small in iodimetric and iodometric titrations of
O·IN solutions; with more dilute solutions, it must be deter-
mined in a liquid having the same composition as the solution
titrated has at the end point.
Starch solution cannot be employed in alcoholic solution,
nor can it .be used in a strongly acid medium because hydro-
lysis of the starch occurs.
Procedure. Weigh out 25 g. of A.R. sodium thiosulphate
Na 2S20 a,5H 20, dissolve it in boiled-out distilled water, and
make up to 1 litre with the same water. Add 3·8 g. of borax
or 0·1 g. of. anhydrous sodium carbonate to stabilise the solu-
tion.. Shake well.
m, 102. Standardisation of sodium thiosulpbate solutions.
-A. With potassium iodate. A.R. potassium iodate has a
purity of at least 99·9 per cent: it can be dried at 120°C.
This reacts with potassium iodide in acid solution to liberate
iodine:
KIO a + 5KI + 3H 2SO, = 3K zSO, + 31 + 2 3H2 0.
Its equivalent' weight as an oxidising agent is 1/6 mol or
214'02/6; a O·IN solution therefore contains 3·567 g. of
potassium iodate per litre.
Weigh out accurately 0·14 to 0·15 g. of pure dry potassium
iodate, dissolve it in 25 mI. of cold, boiled-out distilled water,
it Alternatively, the addition of 0'1 g. of thymol to 100 ml. of boiling water
before mixing with 1 g. of starch produces a starch solution which keep& for
several months [Saville, 1941J.
408 \(uantitative Inorganic Analysis
add 2 g. of iodate-free potassium iodide* ·and 5 m!. of 2N
sulphuric acid (1). Tjtrate the liberated iodine against the
thiosulphate solution with constant stirring. When the
colour of the liquid has become a pale yellow, dilute to ca.
200 ml. with distilled water, add 2 ml. of starch solution, and
continue the titration until the colour changes from blue to
colourless. Repeat with two other similar portions of
potassium iodate.
Calculate the normality of the sodium thiosulphate solution.
1 MI. N Na:2S 20 3 == 0·03567 g. KI0 3 •
Note. 1. Potassium iodate has a small equivalent weight (35·67)
so that the error in weighing 0·14 to 0·15 g. may be appreciable.
In this case it is better to weigh out accurately 3·567 g. of the A.R.
sall: (it. a slightly different weight is used, the exact normality is
calculated), dissolve it in water, and make up to 1 litre in a volu-
metric flask. 25 Ml. of this solution are treated with excess of pure
potassium iodide (1 g. of the solid or 10 ml. of 10 per cent solution),
followed by 3 ml. of 2N sulphuric ' acid, and the liberated iolline is .
titrated as detailed above. The titrations are repeated until two
agree within 0-1 ml.; not more than three titrations s;hould be
necessary. The normality of the sodium thiosulphate solution is
then calculated" (see Section In, 4A).
B. With potassium bromate. Potassium bromate, is very
similar in properties to potassium iodate; its only advantage
over the latter is that it is less expensive. A.R. potassium
bromate has a purity of at least 99·9 per cent. It may be
dried at 120 0 to 150°C, and its aqueous solution is stable.
The substance reacts with potassium iodide in acid solution
in accordance with the equation:
KBr0 3 + 6KI + 6HCl = KBr + 6KCl + 31 2 + 3H 20.
The equivalent weight as an oxidising agent is 1/6 mol; a
O·IN solution therefore contains 167'02/60 or 2·784 grams of
potassium bromate per litre.
Weigh out accurately 0·11 to 0·12 g. of pure dry potassium
bromate (1). Dissolve it in 35 m!. of boiled-out distilled
water, add 2 g. of iodate-free potassium iodide, and 5 m!. of
4N hydrochloric acid (2). Titrate with the sodium thiosul-
phate solution until most of the liberated iodine has reacted,
dilute with 150 m!.. of distilled water, then add 2 m!. of
freshly prepared starch solution, and complete the titration.
'" The absence of iodate is indicated by adding dilute sulphuric acid when no
immediate yellow colouTation should be obtained. If staTch is added, no
immediate blue colouration should be produced.
Repeat the determination with two further similar samples of
potassium bromate.
1 Ml. N Na 2 S20 3 == 0·02784 g. KBr0 3 •
Notes. 1. It is better to prepare a O·1N solution by weighing out
2·784 g. of A.R. potassium bromate, dissolving it in water, and
making up to 1 litre in a volumetric flask (compare Note I under A).
25 Ml. of ·this solution are treated with excess of pure potassium
iodide (1·5 g. of the solid or 15 ml. of the 10 per cent solution),
followed by 4 to 5 mI. of 4N hydrochloric acid, and the liberated
iodine titrated as detailed above. The titrations are repeated until
two agree within 0·1 ml.; not more than three titrations should be
necessary. The normality of the sodium thiosulphate solution is
then calculated.
2. A higher acidity is required than for potassium iodate solutions.
The reaction is somewhat slower; the addition of a few drops of 3
per cent ammonium molybdate solution considerably increases the
velocity of the reaction.
C. With potassium permanganate. This method is satis-
factory provided the experimental conditions given below are
strictly followed. Upon adding potassium permanganate to
an acid solution containing potassium iodide, the following
reaction occurs:
2KMn0 4 + lOKI + 16HCl = 2KCI + 2MnCI 2 + 512 + 8H2 0.
Measure o-qt 25 ml. of standard 0·1N potassium perman-
ganate into a 350 mI. conical, preferably glass-stoppered,
flask containing a solution of 6 g. of potassium iodide in 50 ml.
of water and 2 ml. of concentrated hydrochloric acid. Allow
to stand in the dark for 10 minutes. Titrate the iodine that
has been set free with the sodium thiosulphate solution until
the solution has a faint straw colour, add 2 mr of the starch
s6lution, and continue the titration until the blue colour is
just destroyed. Carry out a blank determination, but
substitute 25 mI. of water for the 25 m!. of the potassium
permanganate solution; subtract the volume of thiosulphate
solution required in the blank run from that consumed in the
other titration. Repeat the determination with two other 25
ml. portions of the permanganate solution.
1 Ml. N Na 2 SZ0 3 == 0·03161 g. KMn0 4 •
D. With potassium dichromate. Potassium dichromate is
reduced by an acid solution of potassium iodide to a green
\
chromic salt, and an equivalent weight'of iodine is set free:
K2Cr207 + 6KI + 14HCI = 8KCI + 2CrCl 3 + 31 2 + 7H 20.
This reaction is subject to a number of errors: (1) it is not
instantaneous, and (2) the hydriodio acid (from excess of
iodide and acid) is readily oxidised by air, especially in the
presence of chromic salts. It is accordingly best to pass a
current of carbon dioxide through the reaction flask before
and during the titration (a more convenient but less efficIent
method is to add some solid sodium bicarbonate to the acid
solution, and to keep the flask covered as much as possible),
and to allow 5 minutes for its completion.
Place 100 ml. of cold, recently-boiled distilled water in a
500 ml. conical, preferably glass-stoppered, flask, add.3 g.
of iodate-free potassium iodide and 2 g. of pure sodium bicar-
bonate, and shake until the salts dissolve. Add 6 ml. of
concentrated hydrochloric acid slowly whilst gently rotating
the flask in order to mix the liquids; run in 25 ml. of standard
O'IN potassium dichromate (1), mix the solutions well, and
wasQ the sides of the flask with a little boiled-out water from
the wash bottle. Stopper the flask (or cover it with a small
watch glass), and allow to stand in the dark for 5 minutes in
order to complete the reaction. Rinse the stopper or watch
glass, and dilute the solution with 300 ml. of cold, boiled-out
water. Titrate the liberated iodine against' the sodium
thiosulphate solution contained in a burette, whilst constantly
rotating the liquid so as to thoroughly mix the solutions.'
When most of the iodine has reacted as indicated by the
solution acquiring a yellowish-green colour, ?-dd 2 m!. of
starch solution and rinse down the sides of the flask; the
colour should change to a shade of blue. Continue the
addition of the thiosulphate solution dropwise, and swirling
the liquid constantly, until I drop changes the colour from
greenish-blue to light green. The end point is sharp, and is
readily observed in a good light agains1: a white background.
Carry out a blank determination, substituting distilled water
for the potassium dichromate solution; if the potassium
iodide is iodate-free, this should be negligible. Repeat with
two other 25 ml. portions of the dichromate solution.
Calculate the normality of the sodium thiosulphate solution'.
1 Ml. N Na 2S2 0 3 == 0·04904 g. K 2Cr 2 0 7 •
Note. 1. If preferred, about 0·2 g. of A.R. potassium dichromate
may be accurately weighed out, dissolved in 50 ml. of cold, boiled-
c:>ut water, and the titration carried out as detailed above.
Volumetric A nalys£s 411
E. With copper. This method is not recommended ~xcept
where the thiosulphate solution is to be used for the deter-
- mination of copper. A.R. copper turnings (prepared from
electrolytic copper) has a purity of over 99·9 per cent. *
Weigh out accurately about 0·25 g. of pure copper turnings
into a 250 to 350 m!. conical flask, and dissolve it in a mixture
of 3 m!. of concentrated nitric acid and 6 m!. of water. Dilute
to 21) m!. with water, heat the solution to boiling, add 0·5·g.
of urea and boil the solution for 1 to 2 minutes. The urea
completely eliminates nitrous acid and oxides of nitrogen,
which interfere with the determination.t Cool the solution,
and add 6N ammonium hydroxide solution slowly until a
slight permanent precipitate of cupric hydroxide just forms.
Add I) ml. of glacial acetic acid to the almost neutral solution.
Cool the solution, and add 3 g. of potassium iodide dissolved
in 30 m!. of water (or 30 m!. of a 10 per cent solution), cover
the conical flask with a watch glass, and allow to stand for 1
minute. Titrate the liberated iodine with the thiosulphate
solution, added from a burette, until the brown colour due to
iodine has changed to light yellow. Add 2 mI. of starch
solution, and continue the titration with the thiosulphate,
stirring constantly, until 1 drop changes the colour from blue
to yellowish-white.+ The cuprous iodide is always coloured
by adsorbed, " starch iodide," and is not a pure white at the
end point. Repeat the determination with two other 0·25 g.
portions of copper.
When starch alone is employed as indicator, the iodine is
not completely available for reaction with the sodium thio-
sulphate and hence the strength of the thiosulphate solution
determined by mea"Qs of copper is slightly higher (ca. 3 per
cent) than that standardised against potassium iodate or
iodine. It is for this reason that the method was not recom-
mended above except where the thiosulphate solution is
* The Bureau of Analysed Samples Ltd., of Middlesbrough, England, supply
copper turnings of 99·91 per certt purity (a " British Chemical Standard ").
t Nitrite and nitric oxide when present even in minute quantities interfere.
Thus nitrite in acip solution reacts in accordance with the equation:
::!NO.- + 21- + 4H+ = 2NO + I. + 2HsO.
The nitric oxide is oxidised by air to nitrogen peroxide, which reacts with
the iodide to form more nitric oxide and iodine; the cycle is then repeated.
For this reason, if a trace of nitric oxide is present in the solution, no perma-
nent end point is obtained upon titration with thiosulphate solution in the
presence of starch. The colour returns rapidly upon the removal of the iodine.
: The end point is rendered sharper by adding 1 or 2 drops of O·lN silver'
nitrate; the yellowish colour of silver iodide neutralises the purplish tinge of
cuprous iodide.
subsequently to be used for the determination of copper.
H. W. Foote (1938) has shown that the iodine adsorbed
by the cuprous iodide is liberated and made available to the
thiosulphate if a soluble thiocyanate is added to the solution
just before the end point is reached in the usual titration.
Under these conditions, the total iodine liberated is closely
equivalent to the copper and the end point is sharp, the pre-
cipitate turning almost white. This simple modification
permits the use of pure copper in the standardisation of thio-
sulphate solutions for general use. The experimental pro-
cedure is exactly as above. The thiosulphate solution is
added until the yellow-brown colour due to iodine is nearly
discharged, 2 ml. of starch solution is introduced, and titra-
tion is continued until near the end poinf. About 1·5 to 2'0 g.
of A.R. ammonium thiocyanate is then added (to liberate
'the adsorbed iodine) and titration is continued until the blue
colour is just discharged.
The principal reaction is :
2Cu++ + 4r ~ CU212 +1 2,
It proceeds from left to right because of the insolubility of the

cuprous iodide under the experimental conditions employed.
1 Ml. N Na2S20a :..:: 0·0357 g. Cu.
Note. J. ]. Kolb (1939) has introduced the following modification
which utilises perchloric acid for solution of the copper. The
procedure has the advantage that all by-products are rapidly and
completely removed by volatilisation (contrast the process for
removing oxides of nitrogen described above) whilst cold dilute
perchloric acid does not interfere in the titration.
Weigh out accurately 0·6 to 0·7 g. of pure copper into a small
beaker provided with a clock-glass or Fisher" speedyvap " beaker
cover, add 6 to 8 ml. of 60 to 70 per cent,perchloric acid· and heat
to boiling in a fume cupboard. Boil gently until solution is com-
plete, continue the boiling for a few minutes, and cool slightly. Add
an equal volume of distilled water, and boil for 2 minutes to elimin-
ate any chlorine which may still be present. Cool to room tempera-
ture, transfer quantitatively to a 100 ml. volumetric flask, and make
up to the mark. Mix well. To '25 m!. of this solution add 30 m].
of 10 per cent potassium iodide solution, allow to stand for 2 min-
utes, and titrate against the thiosulphate solution until the yellow
colour is nearly discharged. Add 2 to 3 m!. of starch solution and.
titrate to near the end point. Add 1·5 to 2 g. of A.R. potassium or
• See note as to danger attending the use of perchloric acid in Section IV. 70.
Volumetric A naly sis 413
ammonium thiocyanate and continue the titration until the blue
colour just disappears. The mixture at the end point is flesh-
coloured.
The results by this method are in close agreement with those
obtained with potassium iodate.
F. With pure iodine. Pure iodine is prepared as described
in Section U, 12B under Sublimation. This method has the
advantage of giving a direct value, but suffers from the draw-
backs that it is necessary to prepare pure dry iodine and the
difficulty of weighing out accurately the volatile iodine.
The weighing of the iodine is usually carried out in some form
of weighing bottle containing a concentrated solution of
potassium iodide; the solution of iodine in potassium iodide
has a lower vapour pressure, thus reducing the volatility.
Select two or three small weighing bottles fitted with well-
ground stoppers and of such size that they will pass completely
through the neck of a 350 or 500 m!. conical flask. Into one
'of these weigh out '2 to 2·5 g. of iodate-free potassium iodide
and 0·5 ml. of water; stopper the weighing bottle and weigh
accurately. Remove the stopper, add 0·4 to 0·5 g. of pure
iodine, quickly replace the stopper, and weigh again; the
difference between the two weighings gives the weight of
iodine. Place the weighing bottle in the neck of the conical
flask held in an inclined position and containing 200 ml. of
water and 'about 1 g. of pure potassium iodide. Allow the
bottle to slide to the bottom of the flask, but just as it falls
remove th~ stopper and allow it to fall also. No iodine is
thus lost. If the contents of the weighing bottle are washed
into the water, a slight loss of iodine may occur. Titrate the
solution, containing a known weight of iodine, with the sodium
thiosulphate solution from a burette. When the liquid is
pale yellow in colour, add 2 ml. of starch solution, and con-
tinue the titration ~ropwise until the bl"4e colour is just
destroyed. Repeat the determination with two other similar
weighing bottles and 0·4 to 0·5 g. of iodine.
1 Ml. N Na 2S20 3 == 0·12692 g. 1.
G. With a standard solution of iodine. If a solution of

iodine which has been standardised against arsenious oxide
(Section m, IOU) is available, this may be employed in the
standardisation of a sodium thiosulphate solution. The
procedure is 'identical with that given under F above.
Calculate the normality of the sodium thiosulphate solution
as based upon that of the iodine solution (compare Section
IU,4A).
m. 103. Preparation of O'lN iodine soIution.-Discussion.

0·335 Gram of iodine dissolves in I litre of water at 25°C.
In addition to this small solubility, aqueous solutions of
iodine have an appreciable vapour pressure of iodine, and
therefore decrease slightly in concentration on account of
volatilisation when handled. Both difficulties are overcome
by dissolving the iodine ·in an aqueous solution of potassium
iodide. Iodine dissolves readily in aqueous potassium iodide,
the more concentrated the solution, t4e greater is the solu-
bility of the iodine. The increased solubility is due to the
formation of a tri-iodide ion:
12 + r- ~ I a-.
The resultant solution has a much lower vapour pressure than
a solution of iodine in pure water, consequently the loss by
volatilisation is considerably diminished. Nevertheless the
vapour pressure is still appreciable so that precautions should
always be taken to keep vessels containing iodine closed except
during the actual titrations. When an iodide solution of iodine
is titrated with a reductant, the free. iodine reacts with the
reducing agent, this displaces the equilibrium to the left, and
eventually all the tri-iodide is decomposed; the solution
therefore behaves as though it were a solution of free iodine.
Iodine solution attacks rubber, hence.it should always be
used with burettes fitted with glass taps. Since iodine and
iodides are comparatively expensive, all residual liquids or
precipitates containing iodine in any form should be returned to the
bottle marked "Iodine Residues."
For the preparation of standard iodine solutions, A.R. or
resublimed iodine and iodate-free (e.g. A:R.) potassium iodide
should be employed. The solution may be standardised
against pure arsenious oxide or against a sodium thiosulphate
solution which has been recently standardised against potas-
sium iodate, potassium bromate, or, less desirably, potassium
permanganate or potassium dichromate. For work of the
highest accuracy purposely-purified iodine may be weighed
out directly, but in view of the volatility of the solid, this
method is not recommended for beginners nor, indeed; is it
necessary for most ordinary purposes. It is given here "for
the sake of completeness.
The equation:
12 + 2£ ~ 21-
indicates that the equivalent weight is equal to the atomic
weight, or 126·92. *
Procedure. Preparation of O'IN iodine.
Method A. Dissolve 20 g. of iodate-free potassium iodide
(e.g., A.R.) in 30 to 40 ml. of water in a glass-stoppered 1 litre
flask. Weigh out about 12·7 g. of A.R. or resublimed iodine
on a watch glass on a rough balance (never on an analytical
balance on account of the iodine vapour). and transfer it by
means of a small dry funnel into the concentrated potassium
iodide solution. Insert the glass stopper into the flask, and
shake in the cold until all the iodine has dissolved. Allow
the.solution to acquire room,temperature, and make up to the
mark with distilled water,
The iodine solution is best preserved in small glass-stoppered
bottles. These should be filled completely and kept in a cool
aark place.
Method B. Purify resubIimed or A,R. iodine as described in
Section IT, 12B. Secure a large weighing bottle provided with
a well-fitting ground glass stopper. Place 20 g. of pure
potassium iodide and 5 ,mI. ·of water into it; the solid will
dissolve' completely on gentle shaking. When the bottle
and contents have acquired room temperature, weigh them
accurately to 0·1 mg. Weigh out 6·0 to 6'5 g. of the purified
iodine on a rough balance and introduce it carefully without
splashing into the weighing bottle. Allow to stand (say, for
20 minutes) in order to attain the room temperature; and
weigh again to the nearest 0'1 mg. Immediately after the
weighing transfer the solution quantitatively to a glass-
stoppered 500 ml. volumetric flask, and make up to the mark
with distilled water. Stopper the flask,. and store it in a
dark place. .
Calculate the normality of the iodine solution from the
weight of pure iodine employed. A O·IN solution contains
12·692 g. of iodine per litre,
m, 104. Standardisation of iodine solutions.t-
A. With arsenious oxide. Discussion. A.R. Arsenious
oxide of 99·9 per cent purity can be obtained commercially;
the U.S. Bureau of Standards supply a product of 99·98 per
• This also follows from the hypothetical equation:
I. + H.O = 2HI + o.
t Prepared by Method A. Section m, 103.
cent guaranteed purity. * This s"4bstance therefore provides
an excellent primary standard for the standardisation of
iodine solutions. The reaction between arsenious oxide and
iodine is a reversible one :
HsAsOa + 12 + H 20 ~ HsAs04 + 2H+ + 2rt,
and proceeds quantitatively from left to right if the hydrogen
iodide is removed from the solution as fast as it is formed.
This may be done by the addition of sodi'um bicarbonate;
sodium carbonate and sodium hydroxide cannot be used since
they react with the iodine forming iodide, hypoiodite and
iodate. Actually it has been shown that the complete
oxidation of the arsenite occurs when the pH of the solution
lies between 4 and 9, the best value being 6·5, which is very
close to the neutral point. Buffer solutions (Section I, 20)
are employed to'maintain the correct pH. A 0'12N solution
of s(lnium bicarbonate saturated with carbon dioxide (car-
bonic acid) has a pH of 7; a solJ,ltion saturated with both
borax and boric acid has a PH of about 6·2, whilst a
Na 2 HPO c NaH 2 P0 4 solution is almost neutral. Any of
these three buffer solutions is suitable, but the first-named is
generally employed as already stated.
Procedure. Weigh out accurately about 2·5 g. of finely
powdered A.R. arsenious oxide (1)., transfer fo a 400 ml.
beaker, and dissolve it in a concentrated solution of sodium
hydroxide, prepared from 2 g. 'of iron-free sodium hydroxide
(e.g., A.R.) and 20 ml. of water. Dilute to about 200 mI.,
and neutralise the solution with N hydrochloric acid, using a
strip of litmus paper as indicator. When the solution is
faintly acid, remove the litmus paper by means of a stirring
rod and carefully rinse both the rod and the paper. Transfer
the contents of the beaker quantitatively to a 500 m!. volu-
metric flask, add 2·g. of pure sodi1!m bicarbonate, and, when
• Arsenious oxide of 99·95 per cent assay value is obtainable from the
Mallinckrodt Chemical Works. U.S.A. .(See Appendix. Sections A, 8, 9.)
If there is any doubt as to the purity of the arsenious oxide. it may be purified
by dissolving in hot dilute hydrochloric acid (1 : 2). filtering and cooling the
filtrate. The supernatant liquor is poured off, the crystals washed with water,
dried at 105° to 110°C for 1 to 2 hours. and then sublimed (Section II, 12B).
The sublimate may be dried for 12 to 24 hours in a desiccator over concentrated
sulphuric acid.
Pure arsenious oxide is not appreciably hygroscopic and. except for work of
the highest accuracy, it is unnecessary to dry it b~fore use.
t The reaction is sometimes written:
AszO a + 2I z + 2HP ~ Asp, + 4HI.
This equation shows that the equivalent weight of arsenious oxide in this
reaction is 1/4 mol, or 197'82/4 = 49·46.
Volumetric
,
4. nalysis 417
,
all the salt has dissolved, dilute to the mark. Shake' well (2).
Measure out from a burette (this is advisable owing to the
poisonous properties of the solution) 25 ml. of the arsenite
solution into a 250 rol. conical flask (3), add 25 to 50 rol. of
water, 5 g. of sodium bicarbonate and 2 ml. of starch solution.
Swirl the solution carefully until the bicarbonate has dis-
solved. Then titrate slowly with the iodine solution, con-
tained in a burette, to the first blue colour.
Alternatively, the arsenite solution may be placed in the
burette, and titrated against 25 mI. of the iodine solution
contained in a conical flask. When the solution has a pale
yellow colour, add 2 ml. of starch solution, and continue the
titration slowly until the blue colour,is just destroyed.
For either method, further tit rations should be carried out
until individual tit rations agree within 0·1 ml. Not more than
3 titrations should be necessary.
Calculate the normality of t~ iodine solution. A O'IN
solution of arsenious oxide contains 4·946 g, As 2 0 a per litre.
Notes. 1. If an exactly O'IN solution is required, this is obtained
by weighing out exactly 2·473 g. of arsenious oxide.
2. For the preparation of sodium arsenite solution with a
Na2HP04 /NaH aP04 buffer mixture, proceed as follows (Washburn,
1908). Weigh out accurately about 2·47 g. of pure dry arsenious
oxide into a 500 ml. volumetric flisk, dissolve it in a concentrated
solution containing 6 g. of pure sodium hydroxide, add 5 m!. of pure
syrupy phosphoric acid (0·15 mot), and dilute to the mark. This
solution will preserve its titration value indefinitely.
3. If it is desired to base the standardisation directly upon arseni-
ous oxide without the intermediate use of a volumetric flask,
proceed as follows. Weigh out accurately about 0·2 g. of pure arseni-
ous oxide into a conical flask, dissolve it in 10 m!. of N sodium
hydroxide, and add a small excess of dilute sulphuric acid (say, 12
to 15 ml. of N acid). Then add carefully a solution of 5 g. of sodium
bicarbonate in 50 ml. of water, followed by 2 mI. of starch solution.
Titrate slowly against the iodine solution to the first blue colour.
Repeat with two other similar quantities of arsenious oxide.
Calculate the normality of the iodine solution.
B. With standard sodium thiosulphate solution. Sodium
thiosulphate solution, which has been recently standardised,
preferl}bly against pure potassium iodate, is employed.
Transfer 25 ml. of the iodine solution to a 250 ml. conical
flask, dilu_te to 100 ml., and add the standard thiosulphate
solution from a burette until the solution has a pale yellow
colour. Add 2 m!. ·of starch solution, and continue the
p
418 Quantitative; Inorganic Analysis
addition of the thiosulp~ate solution slowly until the solution
is just colourless. Repeat the titration with two further 25
m!. portions of the iodine solution. The titrations should
agree wi thin 0·1 ml.
From the known normality of the s~dium thiosulphate
solution, calculate the normality of the iodine solution
(compare Section m, 4A).
ANALYSES INVOLVI~G THE USE OF STANDARD

SODIUM TH10SULPHATE SOLUTION, STANDARD
IODINE SQLUTION" OR STANDARD SODIUM
ARSENITE .SOLUTION
m, 105. Determination of copper in crystallised copper
sulphate.-'-For practice in this estimation, the student may
determine the percentage of copper in A.R. copper sulphat~,
CuSO,.5H 20. -
Method .1. Weigh out accurately about 3·0 g. of the salt,
dissolve it in water, and make up to 250 ml .. in a volumetric
flask. Shake welt Pipette 50 ml. of this solution into a 250
m!. conical flask, add 3 g. of potassium iodide (or 30 m!. of a
10 per cent solution) (1), and titrate th~ liberated iodine with
standard 0 ·IN sodium thiosulphate (Section m, 102,
Procedure E) (2). Repeat the titration with two other 50 m!.
portions of the copper sulphate solution.
The reaction is :
2CuSO, + 4K1 = Cu 212 + 12 + 2K 2SO"
from which it follows that:
2CuSO, == 12 == 2Na 2 S20 a,
or 1 ml. N NazSzO a == 0·06357 g. Cu.
Calculate the percentage of copper in crystallised copper
sulphate.
Notes. 1. If in a similar determination, free 'mineral acid is
present. a few drops of dilute sodium carbonate solution must be
added until a faint permanent precipitate remains, and this is re-
moved by means of a drop or two of acetic acid. The potassium
iodide is then added and the titration continued. For accurate
results, the. solution should have a p:fl of 4 to 5'·5.
2. If difficulty is experienced. in determining the end point, a
large excess (7 to 8 g.) of potassium iodide may 'be added; this
largely dissolves the precipitated cuprous iodide with the formation
of a complex salt. This' modification is .expensive. A simpler
Volumetric Analysis 4.19
method is to add 2 g. of ammonium thiocyanate towards the end of-
the tifration, viz., just after the blue starch colour diminishes in
intensity.
Method B. Discussion. The following method, due to G. Bruhns
(1918). replaces part of the expensive potassium iodide by the cheaper
potassium thiocyanate. It is based upon the fact that cuprous
thiocyanate is 'less soluble than cuprous iodide leading to the
following reaction in aqueous solution:
CU zI 2 + 2KCNS = Cu 2(CNS)2 + 2KI (i),
and also upon the following reactions:
2CuSO, + 4:KI = Cu 212 + 12 +2K 2SO, (ii) ;
12 + 2Na2S20 a = 2Nal + Na 2S,,06 (iii)';
2CuSO, + 4KCNS = 2Cu(CNS)2 + 2K 2SO, (iv) ;-
2Cu(CNS)2 + 2Na2S20 a = Cu 2 (CNS)2 + 2NaCNS + Na 2S"Oe (v).
Bruhns uses two alternative procedures for the titration of a solution
of a copper salt slightly acidified with sulphuric acid. In the first
he employs a solution containing 24·8 g. crystallised sodium
thiosulphate and 25 t~ 30 g. of potassium thiocyanate per litre,
and titrates this with the copper solution in the presence of a
little starch as indicator. In the second procedure he employs
a solution containing 24·8 g. of crystallised sodium thiosulphate,
25 g. of potassium thiocyanate and 8 g. of potassium iodide per litre,
and uses this directly in the titration of the copper solution in the
presence of starch. Only' mercury and silver, which form insoluble
iodides, interfere; the influence of ferric salts is prevented by the
addition of oxalic acid or of a soluble fluoride.
For many purposes it is sufficient to add a little potassium iodide
to the solution, followed by excess of potassium thiocyanate; the
solution is then titrated against standard thiosulphate in the usual
way.
Procedure. Prepare the copper sulphate solution as in
Method A. Pipette 50 ml. of the solution into a conical flask,
add 3 to 5 ml. of 10 per cent potassium iodide solution fol-
lowed by. about 2 g. of potassium thiocyanate. Titrate
immediately against standard O'lN sodium thiosulphate;
towards the end of the titration dilute with about 100 ml.
of water and add 2 ml. of s.tarch indicator. .
Calculate the percentage of copper in crystallised copper
sulphate.
m, 106. Determination of copper in an ore.-Discuss~on.

Of the common elements which ·are usually associated with.
copper ores, those that interfere with the iodometdr. estima-
tion are iron, arsenic and antimony. Ferric iron is reduced
by iodide: •
2Fe+++ + 2r ~ 2Fe++ + 12 ,
but by the addition of .excess of fluoride, the ferric iron is
converted into the complex rFeFar - -, which yields so small
a concentration of ferric ions that it has no oxidising action
upon the iodide. Arsenic and antimony in the trivalent form
react with iodine, but in consequence of the oxidising medium
usually employed to bring the sample into solution they ~ill
be present in the quinquevalent form. Arsenic and anti-
monic compounds will not oxidise iodide in a solution having
a pH greater than about 3·2. By the use of excess of
ammonium bi-fluoride, NH 4HF 2 , which acts as a buffer, the
pH of the solution can be maintained above 3·2; under
these conditions the reduction of the cupric ion proceeds to
completion. The concentration of the fluoride should be 1·.0
to.l·6 molar.
Procedure. Weigh out accurately 0·5 to 1·0 g. of the finely
powdered copper ore* into a 250 m!. beaker, add 10 to 15
ml. of concentrated nitric acid, and heat until the ore is well
disintegrated and all the copper is in solution. Evaporate to
a volume of 5 m!. (e.g., on an electric hot plate). Add 10 ml.-
of concentrated hydrochloric acid and 10 ml. of I : I sulphuric
acid, and evaporate to dense white fumes. To complete the
oxidation of arsenic, etc., add cautiously 20 ml. of water and
10 ml. of saturated bromine water,.. and boil the solution until
all the btomine is expelled. Filter off the residue on a small
quantitative filter paper (1), and collect the filtrdte in a 250
mI. conical flask. Wash the residue with hot I per cent
nitric acid until the washings are colourless. Concentrate the
nitrate to about 30 ml. (place a short funnel in the neck of the
flask), add 0·5 g. of urea to' the hot solution, and heat to
boiling for I minute; this removes any nitrous acid or oxides
of nitn;>gen which play be present. To the cold solution add
ammonia solution (1 : 1) dropwise until the solution smells
slightly of ammonia. This will preeipitate all the iron as
hydroxide; an excess of ammonia solution should be avoided.
Add 2·0 g. of ammonium bi-fluoride NH4HF2 (2) and shake
until all the ferric hydroxide has .qissolved. Now add 3 g.
of A.R. potassium iodide dissolved in 5 to 10 ml. of water, and
titrate at once with standard O'IN sOQ_ium' thiosulphate,
• Ridsdale's .. Copper pyrites, N,o. 26" (one of the Analysed Samples for
Students) is suitable.
Volume'tric Analysis 42i
which has been standardised against pure copper (Section
m. 102, Procedure F), adding 2 rol. of starch solution near the
end' point. The colour change is from blue to yellowish-
white; the blue colour should not return within 10 to 15
minutes (3). Repeat the determination with another sample
of the ore.
Calculate the percentage of copper in the copper ore.
Notes, 1. If the residue is small In amount or light coloured, the
filtratiOn may be omitted. The subsequent washing with 1 per
cent I}itric acid and treatment with urea is then unnecessary.
2. About 1 'g. of ammonium bi-fluoride should be added for each
0,1 g. of iron present.
3. If Foote's thiocyanate method (Section m, \).02E) is employed,
the colour change is more or less permanent.
m, 107. Determination of manganese dioxide in pyrolusite.
-Discussion. All higher oxides of the heavy metals which
evolve chlorine upon treatment with conyentrated hydro-
chloric acid may be determined with accuracy by passing the
chlorine into excess of potassium iodide. solution, and titrating
the liberated iodine with standard sodium thiosulphate or
sodium arsenite solution. It is essential to be certain that all
the chlorine reacts with the pota;ssiurn iodide solution.
With pyrolusite, the following reactions occur:
Mn0 2 + 4HCI = MnCI I + 2H 20 + cia;
2KI + C1 2 = 2KCI + 12 ,
Procedure. A distillation apparatus, such as is portrayed
diagrammatically in Fig. 3-8 or Fig. 3-9, is employed. The
apparatus fitted with ground glass joints is to be preferred
owing to the attack of co _
chlorine upon rubber. 2
The apparatus in Fig. 3-8

consists of a distillation flask
of about 200 mI. capacity,
connected to two U (Peligot)
tubl!s containing potassium
iodide solution. In Fig. 3'-9
a ~50 to ?OO mt flask is pro-
vided with two ground glass
joints at A and B; the tube -'~~ ,~',,'
from B passes into a small ~-~ "-
conical flask containing pot- :::?:.:.~::.':::0."-
assium iodide solution; C is d====5t~~~,,~~,~::::--":::::2'~.,,~\
a small tube containing glass FIG. 3-8.
beads moistened with potassium iodide solution; the rest
of the apparatus is self-explanatory. Weigh out accurately
about 0'2 g. of the finely
powdered mineral * into a
small weighing tube (1 cm.
wide and 2 cm. long is a con-
vt:nient size), and introduce the
tube and contents into the
distillation flask; connect up
the apparatus as shown in the
Figure. Introduce 100 rill. of
water containing 3·5g. of iodate-
free (e.g., A.R.) potassium
iodide into the absorption
apparatus, which should be
immersed in cold water. Place
25 mi. of concentrated hydro-
FIG. 3-9. chloric acid into the separating
funnel, and add it to the contents of the flask ;. meanwhile
maintain a very slow stream of carbon dioxide through the
appar;ltus. Heat the mixture very gently so that_ chlorine
is slowly evolved. Increase the temperature gradually to
the boiling point, and finally pass a slow ct}rrent of carbon
dioxide through the boiling solution until all 'the chlorine has
been expelled (about 15 minutes). Immediately before the
heating is stopped, increase the current of carbon 'dioxide in
order to prevent the solution in the absorption vessels from
passing back into the flask. Disconnect the absorptiQn
vessels. With the apparatus of Fig. 3-8, the contents of the
U-tubes are quantitatively transferred to a conical flask:
with the apparatus of Fig. 3-9, the tubes and be::tds are well
rinsed with water and removed from the flask. Titrate the
liberated iodine at once with standard O'IN sodium thiosul-
phate in the usual manner. If desired, the solution may be
made up to l(j0 m!. in a volumetric flask, and.50 ml. portions
titrated against the sodium thiosulphate solution; if this
procedure is adopted, the weighf of the mineral may be
increased to 0·3 to 0·4 g.
CalcUlate the percentage~of Mn02 in the mineral.
1 Ml: N Na 2S2 0 3 == 0·04347 g. Mn0 2 •
... Ridsdale's .. Pyrolusite (Mn) Ore, No. 18" (one of the Analysed Samples
for Students) or the Bureau of Analysed Samples " Manganese Ore ' A ' "
(one of the British Chemical Standards) is suitable.
Note. Red lead PbaO, and lead dioxide Pb02 may be estimated in
a similar manner (compare Section m, 57, Note).
m, 108. Determination of chlorates and of bromates.-
Discussion. Chlorates and bromates may' be estimated by
the distillation process described in the previous Section for
pyrolusite. The process of distillation may be avoided in
these and similar cases· by mere digestion with hydrochloric
acid and potassium iodide solution at an elevated tempera-
ture; decomposition is quantitative.
KCIO a + 6RCI = KCI + 3CI 2 + 3R 20;
3Cl 2 + 6K1 = 6Kel + 31 2,
Na 2S 2 0 a == I == KCIO a /6 == KBr0 3 /6.
A strong bottle (tr pressure bottle ") with a very accurately
ground stopper is necessary. Generally,
the ordinary stoppered bottle of commerce
is not sufficiently tight, and it is better
to regrind the stopper with a little very
fine emery and water. It must then be
tested by securing the stopper tightly
and immersing in hot water; if any
bubbles of air pass through the stopper,
the' bottle is useless. The stbpper may be
secured by means of fine copper wire'
threaded through a piece of rubber tubing
securely fixed to the neck of the bo,ttle, or
by means of a special screw clamp (Fig.
3-10)·t ~
For practice, the student may determine
the percentage of Cl0 3 (or of Br0 3) in
potassium chlorate (or potassium bromate),
preferably. of A.R. quality. .
Procedure. Weigh out accurately 0·25
toO, 30 g. 0 f the finely-powdered FIG. 3-10,
• For example, chromates and dichromates. These yield the green chromic
chloride and an equivalent quantity of iodine:
2IF.crO, + 6Kl_ + 16HCI = 10KCI + 2CrCI, + 3I t + SHoO ;
K.Cr.0 7 + 6KI + 14HCI = 8KCI + 2CrCl a + 3I t + ·7H,O.
Here it is necessary to dilute the solution very considerably in order to prevent
the green colour of the chromic,salt from masking the blue colour of the starch
in the subsequent titration with standard--sodium thiosulphate solution.
1 MI. N Na,S,Oa == 0,06474 g. K,CrO, == 0·04904 g. K,Cr,01'
t Such bottles may be purchased, for example, from A. Gallenkamp and
Co. Ltd., London, E.C.2, England, or from the Fisher Scientific·Co., Pittsburgh,
Pa., U.S.A.
potassium chlorate into a pressure bottle, add 7 g. of
iodate-free potassium iodide dissolved in a small volume of
water, and displace the air from the bottle by means of
carbon dioxide .from a Kipp's apparatus or from a
cylinder. Add 25 ml. of concentrated hydrochloric acid,
insert the stopper, fasten it down securely, and wrap the
bottJe in a thick duster. Itnmerse the whole in cold water
in a water bath, heat the water gradually to boiling, and
continue the boiling for I hour. A1low the water to cool,
and when quite cold, Temove the bottle, open it, transfer the
contents quantitatively into a 250 ml. volumetric flask, and
make up to the mark. Shake well. Titrate 50 ml. portions
with. standard O·IN sodium thiosulphate, using starch as
indicator. ,.
Calculate the percentage of CIO a in the sample of potassium
chlorate.
I Ml. N Na 2S20 a = 0·1l39rg. ClO a == 0'02043 g. KCIO a.
m, 109. Volumetric determination of lead.-DisCttssion.

The lead is precipitated as the chromate with potassium
chromate or dichromate solution in a medium buffered with
a mixture of acetic acid and sodium or-ammonium acetate.
The precipitate is washed with water, dissolved in diiute
hydrochloric acid, treated with potassium iodide solution, and
the liberated iodine titrated with a standard solution of
sodium thiosulphate.
Pb(CzHaOz)z + K2~r04 = PbCrO, + 2K.C 2 H aOz ;
PbCrO, + 2HCI = H ZCr0 4 + PbCl z ;
2H zCrO, + 6K1 + I2HCl = 2CrCI a + ?KCl + 31 2 + 8HzO~
Thus Pb=!,31, or I ml. N 1=1 ml. N Na ZS20 a =0·06.907 g. Pb.
Another method, which is sometimes used, consists in

adding excess of O'IN potassium dichromate to the' solution
buffered with alkali acetate, hoiling for 2 to 3 minutes,. filter-
ing the precipitate on a sintered J ena glass crucible, washing
the precipitate well with cold water, and determining the
excess of dichromate in the copIed filtrate either iodometric-
ally by-the addition of excess of potassium iodide and titration
with standard thiosulphate solution, or with standard ferrous
ammonium sulphate solution.
For practice, the student may determine the perce~tage of
lead in lead ~itrate, preferably of A.R. purity.
Procedure. Weigh out accurately about 0,3 g. of A.R. lead
nitrate, dissolve it in excess of dilute acetic acid (1 : 4), and
add 2 g. of sodium acetate*; dilute to 150 m!. Heat the
solution to boiling, and add from a pipette 10 ml. of hot 4
per cent potassium chromate solution. Boil gently for 5 to
10 minutes, filter the lead chromate through (' sintered glass
crucible, and wash several times with dilute sodium acetate
solution. Transfer the precipitate quantitatively to a 250
ml. conical flask, dissolve it in dilute hydrochloric acid (1 : I),
add 20 ml. of 10 per cent potassium iodide solution and stir
gently. Titrate at once with standard O·IN sodium tqio-
sulphate until the iodine colour has nearly disappeared, add 2
m!. of starch solution, and continue the titration until the
colour of the solution changes to a clear green. Dilution
facilitates the detection of the end point. ..
Calculate the percentage of lead in the sample of lead
nitrate.
m; 110. AnalySis of hydrogen peroxide.-Discussion. Hy-
drogen peroxide reacts with iodide in aCid solution ih accord-
ance with the equation: •
Ha02- + 2H+ + 2r -: .Ia + 2HaO.
The reaction velocity is comparatively slow, but .increases
with increasing concentration of acid. The addition of 3
drops of a neutral 20 per c~nt ammonium molybdate solution
renders the reaction almost instantaneous, but as it also
accelerates the atmospheric oxidation of the hydriodic acid,
the titr(\,tion is best conducted in an atmosphere of carbon
Clioxide.
The iodometric method has the advantage over the per-
manganate method (Section m, 55) thai it 'is less affected by
stabilisers which are sQIlletimes added to commer<:ial hydrogen
peroxide solutions. These preservatives are often boric
~cid, salicylic acid and glycerol, and render the results ob-
tained by the permanganate procedure less accurate.
Procedure. Dilute the hydrogen peroxide solution ·to 2-
volume strength (ca. 0·6 per cent Ha02)' Thus, if a " 10-
volume" hydrogen peroxide is used, transfer 20 m!. by means
of a burette or pipette to a 100 m!. volum.etric flask, and
make up to the mark. Shake well. Remove 10 ml. 'of this
diluted solution, and add it gradually and with constant
stirring to a s'olufion of 2 g. of pure potassium iodide in 200 m!.
• The addition of alkali acetate is only essential when potassium dichromate
(e.g., 10 ml. of the saturated solution) is used for precipitation.
p*
of 2N sulphuric acid (1 : 20) tontained in a stopper-ed bottle.
Allow the mixture to stand for 15 minutes, and titrate the
liberated iodine with standard O·IN sodium thiosulphate,
adding 2 ml. of starch solution when the colour of the iodine
has been nearly discharged. Run a blank determination at
the same time.
B~tter results are obtained by transferring 10 ml. of the
diluted hydrogen peroxide solution to a conical flask, and
adding 180 ml. of 2N (1 : 2'0) sulphuric acid. Pass a slow
stream of carbon dioxide through the flask, add 20 mI. of 10
per cent potassium iodide solution, followed by 3.drops of 20
per cent ammoniul!l molybdate solution. Titrate the liber-
ated iodine immediately with standard O·IN sodium thio-
sulphate in the usual way.
Repeat the -titration with two other 10 ml. portions of the
peroxide solution. -
Calculate the weight of H 20 2 in 1000 m!. of the original
solution.
IMl. N Na 2SPa =:; ~'01701 g. H 20 2.
Note. The above method is available for all per-salts.
m, 111. Determination of the av8nable chlorine in bleach-
ing powder.-Discussion. Bleaching powder consists essen-
tially of a mixture of calcium hypochlorite Ca(OCl)2 and the
basic chloride CaC1 2,Ca(OH}2,H 20; some free slaked lime is
usuall¥ present. The active cons~ituent is the hypochlorite,
which is responsible for the bleaching action. Upon treating
bleaching powder with acid, chlorine is liberated:
Ca(OC1)2 + 4HCI = CaCl 2 + 2C12 + 2H 20.
The available chlorine refers to the chJoririe liberated by the
action of dilute acids, and is expres~ed a~ the percentage by
weight of the bleaching powder. Th~ bleaching powder of
-commerce cont~ins 36 to 38 per cent of available chlorine.
Two methods are in common use for the determination of
the available chlorine. In the first, the bleaching powder
solution or suspension is treated with an excess of a sotution
of potassium iodide, and strongly acidified with acetic acid:
Ca(OClh + 4KI +. 4H.C2HaOz
= CaCl 2 + 4K.C 2H a0 2 + 212 + 2H 2 0.
The liberated iodine is titrated with standard sodium
thiosulphate solution. The solution should not be acidified
with hydrochloric acid, for the little calcium chlorate
which is usually present, by virtue of the decomposition of the
... hypochlorite, will react slowly with the potas~ium iodide and
lib"erate Iodine:
HCIO a + 6KI + 5HCl = 6KCl + 31 2 + 3H 2 0.
In the second method, the bleaching powder solution or
suspension is titrated against standard O'IN sodium arsenite.
No internal indicator is at present available for the detection
of the end point, since hypochlorite destroys most organic
dyes. An external indicator, potassium iodide-starch p'ilper,
is employed. As long as excess of hypochlorite is present, a
drop of the solution being titrated will turn iodide-starch
paper blue; at the end point, a drop of the solution will no
longer produce a blue stain. The -external indicator may be
dispensed with by adding excess of the standard arsenite
solution, and titrating baGk with standard iodine solution,
using starch as indicator.
Ca(OClh + As 20 3 = CaCl 2 + As 20 5 •
The potassium iodide-starch paper may be purchased, or it can.be
prepared as follows. Boil I g. of starch with 100 m!. of water, filter
the solution, and add 0·1 to 0·2 g. of pure potassium iodide to the
filtrate. Dip strips of filter paper into the solution, and dry them
on a porcelain or glass plate at 40 0 to 50°C.,
Procedure A (iodometric method)~ Weigh out accurately
about 5 g. of the bleaching powder into a clean glass mortar.
Add a little water, and rub the mixture to a smooth paste.
Add a little more water, triturate with the pestle, allow the
mixture to settle, and pour off the milky liquid into a 500 ml.
volumetric flask. Grind the residue with a little more water,
and repeat the operation until the whole of the sample has
been transferred to the flask either in solution or in a state of
very fine suspension, and tlie mortar washed quite -clean.
The flask is then filled to the mark with distille{i water, well
shaken, and 50 ml. of the turbid liquid immediately with-
drawn with a pipette. This is transferred to a 250 ml.
conical flask, 25 ml. of water added, followed by 2 g. of
iodate-free potassium iodide (or 20 ml. of a 10 per cent solu-
tion) and 10 ml. of glacial acetic acid. Titrate the liberated
iodine with standard O'IN sodium thiosulphate; add 2 ml.
of starch solution when the solution has a pale yellow colour,
and continue the titr.ation dropwise until the blu~ colour just
disappears. Repeat the determination with two other 50
ml. portions of the well-mixed suspension.
Calculate the available chlorine in the bleaching powder.
I M1. N Na 2S 2 0 3 =: 0·03546 g. C1.
Procedure B (arsenite method). Prepare a bleaching pow-
der solution or suspension as is described in Procedure' A.
Transfer 50 ml. of the well-mixed suspension to a 250 ml.
beaker, and titrate with the standard O'IN sodium arsenite
(Section III, lO4A) , delivered from a burette, until a drop of
the mixture removed with a glass rod and brought into
contact with potassium iodide-starch paper gives no blue stain.
The first titration will be approximate. In subsequent
titrations, add the arsenite solution until within) rnl. of the
previously determined end point, and remove test drops only
during the completion of the titration.
The use of the external in~ica.tor may be avoided by adding
excess-(say, 75 ml.) of the standard sodium arsenite solution,
and titrating back the excess with standard O'IN iodine,
using starch as indicator. .
Calculate the percentage of available chlorine in the sample
of bleaching powder .
. I Ml. N ASzO a == 0·03546 g. C1.
m, 112. Determination of hypochlorites.-This problem
arises in the analysis of sodium hypochlorite solution and of
"high test hypochlorite" (H.T.H.). The latter substance,
marketed in America, consists of calcium hypochlorite of 75
to 100 per cent purity; it is more stable than ordinary bleach-
ing powder, keeps well, is almost non-hygroscopic, land gives
a practically clear aqueous solution.
Both the iodometric and sodium arsenite methods des-
cribed in the previous Section may be used, and need therefqre
not be repeated here.
III. 113. Determination of antimony.-Discussion. Anti-
monious oxide reacts with iodine in a similar way to arsenious
oxide (Section III. 104). and the antimony may be deter-
mined by titration with a standard solution of iodine:
Sb2 0 a + 2I z + 2H zO ~ Sb z0 6 + 4HI.
The hydriodic acid must be removed as formed by the addition
of excess of sodium bicarbonate solution. To prevent the
precipitation of antimonious hydroxide when an acid solution
is diluted or neutralised, some tartrate is added, which forms
a soluble complex salt, for example:
SbCl a + "KH.C4 H,06 + 3NaHCO a
= K(SbO)C,HP6 + 3NaCI + 2H 20 + 3C0 2 ;
K(SbO)C 4 H,06 + I2 + 4NaHC0 3
= KNa.C,H,06 + NazHSbO, + 2NaI + H 20 + .4CO z.
• Volumetric Analysis 429
This method is satisfactory for tartar emetic, antimony. oxide,
ap.d certain other antimony compounds, but fails when- ap-
plied to alloys' in which antimony is not the principal con-
stituent (5 to 20 per cent antimony). The potassium bromate
method (Sections m, 135, 137) or the permanganate method
(Section m, 117) may then be used, and is, in' general, more
satisfactory for the volumetric estimation of antimony.
antimony in tartar emetic KiSbO)C 4H 4 0 6 ,O'5H 2 0 or in the
anhydrous A.R. potassium antimonyl tartrate.
Procedure (antimony in tartar emetic). Weigh out accur-
ately about 2 g. of the finely-powdered A.R. tartar emetic
(or 2·5 g. of the hydrated compound), dissolve it in water, and
make up to 250 ml. in a volumetric flask. Shake wyll. Trans-
fer 50 ml of the solution to a 250 m!. conical flask by means
of a burette (a pipette should NOT be used, for antimony salts
are nearly as poisonous as those of arsenic), add 20 m!. of a
cold saturated solution of pure sodium bicarbonate (about
10 m!. are required to about 0·1 g. of Sb 20 3 ) and 2 ml. of
starch solution. Titrate with standard O'IN iodine until the
blue colour jus.t appears. The titration must be carried out
immediately after the sodium bicarbonate solution is added,
otherwise a portion of the metal is precipitated as antimoni-
ous hydroxide, which reacts extremely slowly with the iodine.
Repeat the determination with two other 50 ml: volumes of
the solution.
Calculate the percentage of antimony in the sample used.
1 Ml. N I == 0·06088 g. Sb.
m, 114. Determination of antimony in antimonic oxide-

and in antimonates.-Here the reaction is the reverse of that
for the determination of antimonious comp'ounds :
The !lntimonic oxide or antimonate is treated with. concen-

trated hydrochloric acid and an at least five-fold excess of
potassium iodide in an apparat-us similar to that used in
Section 'm, 107, and the iodine liberated is collected in
potassium iodide solution. The free iodine in the distillate
is then tItrated with a standard solution of O'IN sodium
thiosulplia teo
I Ml. N Na 2 S2 0 a == 1 ml. I == 0·08088 g. Sb'JOs'
m t 115. Determination of arsenic in arsenic oxide and in
arsenates.-The reaction is the reverse of that employed in
the standardisation. of iodine with sodium arsenite solution
(Section mt 104A) :
As 20l) + 4H+ + 4r ~ As 20 a + 212 + 2H 2 0.
For the estimation of arsenic in arsenic pentoxide or in,
say, commercial lead arsenate, shake an amount of the
material equivalent to about 0·15 g. AS 20 3 (e.g., 0·5 g. of lead
arsenate) in a glass-stoppered bottle with 30 ml. of concen-
trated hydrochloric acid, add 0·5 g. of potassium iodide, and
after 15 minutes titrate the liberated iodine with standard
O·lN sodium thiosulphate, using starch as indicator. It is
desirable, and for very small quantities of ar _;enates it is
essential, to work in an atmosphere of carbon dioxide in order
to avoid the liberation of iodine by exposure to the air. A
" blank" should be rUJi simultaneously with the iodide and
acid under the same conditions, and the resulting thiosulphate
titre deducted as a correction.
1 Ml. N Na SS2 0 a ~ 1 ml. N I == 0'05746 g. AssO Il •
Note. For those arsenai.~s which do not yield an insoluble salt
with sulphuric acid, 12N sulphuric acid gives good results.
m, 116. Determination of tin.-Discussion. 'The method
depends upon the oxidation of stannous chloride to stannic
chloride by means of iodine solution in the presence of con-
centrated hydrochloric acid:
SnCl 2 + 2HCl + 12 = SnCl,4 + 2HI.
A small amount of sulphuric acid is not objectionable, but
nitric acid must be absent. Very few substances, particu-
larly arsenic and antimony, ordinarily met 'with in analysis
for tin interfere: foreign substances must not be present in
such concentrations that they affect the action of the star('h
indicator. The determination must, however, be conducted
in a non-oxidising atmosphere; this is achieved by main-
taining an atmosphere of carbon dioxide from a Kipp's appara-
tus or from a cylinder. .
If the tin is present in the stannic state, it must be .reduced
by warming with pure (e.g., A.R.) antimony (powdered), iron,
lead, aluminium or nickel. When nickel is employed in the
reduction, a large excess of hydrochloric acid must be present
in order to prevent the precipitation of antimony which may
be re.quired in some analyses, for example, in type metal.
Volumetric Analysis' 4-31
Attention is directed to the fact that it has been shown
(F. L. Okell and ]. Lumsden, 1935) that the iodimetric
. titration of stannous tin is subject to a variable error due to
the presence of dissolved oxygen in the iodine solution. The
error may be as higb. as 1 per cent for O'IN iodine and 10 per
oent. for O'OIN iodine. It is accordingly recommended that
for the determinations of tin described below the iodine
solution be prepared in a carbon dioxide atmosphere with
boiled-out distilled water which has been allowed to cool in
carbon dioxide. The solution should be stored in the ~elf
filling apparatus shown in Fig. 2-25 or Fig. 2-26 under carbon
dioxide, the air in the burette and storage bottle being of
course first displaced by carbon diOXIde. It is' found that
iodine does not appreciably attack red rubber "pressure-
tubing," but black lightly vulcanised tubing is unsatisfactory.
For practice, the student may estimate the tin in, and
hence the purity of, A.R. stannous chloride SnCl 2,2H 2 0, and
also the1:in in an alloy. •
Procedure A (tin in stannous chloride). Weigh out accur-
ately about 2 g. of A.R. stahnous chloride, 'dissolve. it in 150
ml. of concentrated hydrochloric acid, transfer to a 250 ml.
volumetric flask, and make up to .the mark with air-free
distilled water. No appreciable oxidation of the tin occurs
in this process. Transfer 50 ml. of the solution to a 250 !ill.
conical flask, dilute with 50 ml. of air-free water, and add 2 ml.
of starch solution. Pass a current of carbon dioxide over the
surface of the liquid, and titrate with standard O·IN iodine'
to the first appearance of a blue colour. Repeat the titration
with two other 50 ml. portions of the solution. .
Calculate the percentage of tin in the salt used, and there-
from its SnCI 2 ,2H 2 0 content.
1 Ml. N I == 0'05935 g. Sn.
Procedure B (tin in an alloy). Dissolve 0~2 to 2·0 g.,
accurately weighed, of the finely-divided alloy,* according to
the percentage of tin present, in concentrated hydrochloric
acid, transfer the solution fiJuantitatively to a 350 ml. or 1500
mL wide-mouthed conical flask, add 50 ml. of concentrated _
hydrochloric acid, and dilute the solution to 200 ml. Add
about I g. of finely-powdered pure ·antimony, e.g., A.R.
(3SnC1 4 + 2Sb = 3SnCl 2 + 2SbC1 3 ) or 1 g. of A.R. iron wire
I
* Ridsdale's "White Metal, No. 8b" (one of the Analysed Samples for
Students) or the Bureau of Analysed Samples" White Metal: 13 ' .. is suitable.
About 0·2 g. should - -
(SnCl 4 + Fe = FeCl 2 + SnCI 2 ) or I g. of aluminium turnings
(3SnC1 4 + 2AI = 3SnCl 2 + 2AICI 3). Insert a three-hole
rubber stopper, carrying in one hole a gas inlet tube extending
to the bottom of the flask, in the second hole a vertical con-
denser tube 15 to 20 cm. in length, and.in the third a small
stopper. Pass a slow stream of carbon dioxide through the
flask, gradually heat to boiling, and continue the gentle
ebullition for 30 minutes. Cool (preferably in ice) to about
10°C whilst a more rapid stream o~ carbon dioxide is passed
through. When the solution is cold, continue the-ourrent of
the gas, remove the stopper in the third hole, add 2 m!. of
starch solution from a pipette, and insert the tip of a burette
containing standard O·IN iodine. Titrate to the first per-
manent blue colour.
The iodine solution may be standardised against arsenious
oxide or, better, against A.R. tin which has been dissolved
and treated as in the above method.
Calculate the percentase of tin in the alloy.
In, 117. Determination of antimony and tin in type metal.-
Discttssion. The method given below combines the excellent
permariganate method for tb.e determination of antimony:
4Mn0 4 - + 5Sb 20 a + 12H+ = 4Mn++ + 5Sb 2 0 s + 6H 2 0
or' Sb 2 0 a + 20 = Sb 2 0 s ,
and the iodimetric method for tin given in the previous
S~ction.
Procedure. Determination of antimony. Weigh out accur-
ately I g. of the finely-divided alloy* (or an amoun't corres-
ponding to approximately 0·1 g. of antimony), dissolve it in
20 to 50 ml. of concentrated sulphuric acid (according to the
amount of lead present), allow to cool, dilute very carefully
to 100 mI., and allow to stand. Decant the clear liquid from
the lead sulphate formed into a wide-mouthed 350 or 500 m!.
conical flask, and boil for 5 minutes to expel sulphur dioxide.
Boil the lead sulphate with tl_!e appropriate volume of con-
centrated hydrochloric acidt* and an equal volume of water.
When the lead sulphate is completely in solutiop, wash ba'Ck
quantitatively into the main antimony solution, disregarding
• The Bureau of Analysed Samples" White Metal 'A ' " (one of the British
ChefIllcal Standards) is suitable for this estimation. AbQUt 1 g. should be
used.
t Cone. H 2SO" ml. 0 , 10 20-- 30
Cone. 'HCI, m!. 30-50 _ 30-35 15--20 10-15 (W. Pugh, 1933).
These volumes. apply, to a final voh~me of 200 mi.
any lead salt which may now be reprecipitated. Titrate the
solution at 40° to 60°C with standard O·lN potassium per-
manganate, added slowly to prevent loss of chlorine, and
when the end point has almost been reached add a few drops
of methyl orange, and continue the titration untit the colour
of the indicator is discharged (compare Section m, 135).
The potassium permanganate'solution may be standardised
against arsenious oxirle or sodium oxalate in the usual way.
It is better, however, to use finely-powdered A.R. antimony·
or anhydrous A.R. potassium antimonyl tartrate. The
standard antimony solution may be prepared by dissolving
about 3·0 g. of pure antimony, accurately weighed, in 30 mi.
of concentrated sulphuric acid, diluting it, boiling off traces of
sulphur dioxide, and making up to 500 mi. in a volumetric
flask with 125 mI. of concentrated .hydrochloric acid and
water.
In the analysis of lead-tin alloys, the ratio of hydrochloric
to sulphuric acid is important for the attainmer.t of correct
results; ,too little hydrochloric acid leads to the precipitation
'of a basic antimony salt, whilst to~ much may cause reduction
of, the permanganate by chloride ions, with the liberation of
chlorine, and the end point becomes barel{perceptible.
Determination of tin. After the titration of the antimony
by the permanganate solution is complete, idd sufficient
concentrated hydrochloric acid to bring this to 20 per cent
by volume of the entire solution, and add 1 g. of finely-
po.wdered pure antimony or 1 g. of A.R. iron wire. Insert a
three-holed rubber stopper, and proceed exactly as described
for the determination of tin'in Section m, 11GB.
Calculate the percentages of antimony and of tin in the
• alloy. .
m, 118. Determination of sulphurous acid and of sulphites.-
Discussion. The iodimetric estimation is based upon tlie
equation:
H 2S0 3 + H 20 + 12 = H 2S0 4 + 2HI.
For accurate results, the following experimental conditions
must be observed:
(i) the solutions shall be very dilute;
(ii) the sulphite must be added slowly and with constant
stirring to the iodine solution, and not conversely; and
(iii) exposure of the sulphite to the air be minimised.
• A.R. antimony can be obtained from the Mallinckrodt Chemical Works,
U.S.A. See Appendix. Section A, 9,
434. Quantitative Inorganic Analysis
Under these conditions the following possible errors are almost
entirely avoided:
(a) loss of sulphur diox~de by volatilisation;
(b) precipitation of sulphur by the " total" reaction.
3H 2SO S == 2H 2S0 4 +. S + H 20,
which is'tatalysed by hydrio,dic acid, and also by the reaction
H 2SO a + 4HI = 2I2 + S + 3H 2 0 ~ and
(c) oxidation of sulphite to sulphate.
In detf(rminations of sulphurous acid and sulphites, excess of
standard O·IN iodine is diluted with several volumes of water,
acidified with hydrochloric or sulphuric acid, and a known
volume' of the sulphite or sulphurous acid solution is added
slowly and with constant stirring from a burette, with the jet
close to the surface of the liquid. The excess of iodine is then
titrated with standard O·IN sodium thiosulphate. Solid
solusle sulphites are finely powdered and added directly to
the excess of the iodine solution. Insoluble sulphites (e.g.,
calcium sulphite) react very slowly, and must be in a very fine
state of division.
Na 2SOs,7H 2Q in the ordinary or A.R. crystallised product.
Proceclure. Weigh out accurately about 0·25 &". of crys-
tallised sodium sulphite (I), and add it to a 350 ml. conIcal
fillSk containing 25 m!. of standard O'IN iodine, 5 m!. of 2N
hydrochloric acid and 150 m!. of water. Swirl or stir the
liquid until all the solid has reacted, and titrate the excess of
iO,dine against standard O'IN sodium thiosulphate with starch
as indicator. Repeat the determination with two other 0·25
g. portions of the sulphite.
Calculate the percentage of Na 2SO a,7H 20'in the sample
employed,
1 Ml. N I == 0:03203 g. S02 E 0·1261.g. Na25.0a,7H20.
Note. 1. Altema~ively, weigh out accurately about 2·5 g. of
crystallised sodium sulphite, dissolve it in boiled-out water, and
m!l.ke the solution up to 250 ml. in a volumetric flask with boilep.-
out water. Shake well. Use 25 ml. for each titration.
m, 119. Determination of hydrogen sulphide and sulphides.
-Discussion. the iodimetric method utilises the reversible
reaction:
For reasonably satisfactory results, the sulphide solution
must be dilute (concentratio'n not greater than 0'01= per cent
or 0·02N), and the sulphide solution added to excess of
acidified O'OIN or O·IN iodine and not conversely. Loss of
hydrogen sulphide is thus avoid~d, and side reactions are
almost entirely eliminated. (With solutions mor~ concen-
trated than about 0'02N, the precipitated sulphur encloses a
portion of the iodine, and this escapes the subsequent titration
with the standard sodium thiosulphate solution.) The excess
of iodine is then titrated against standard thiosulphate solu-
tion, using starch as indicator.
. I Ml. N I = 0·01703 g. H 2S
Excellent results are obtaine'd by the following method,
which is of wider applicability. When .ex~ess of standard
sodium arsenite solution is treated with hydrogen sulphide
solution and then acidified with hydro'Chloric acid, arsenious
sulphide is precipitated:
aa
ASzO a + 3H zS = As 2S a + 2 0.
The excess of arsenious acid is determined with O'IN iodine
and starch (compare Section m. l04A).
For practice, the student may determine the strength of
hydrogen sulphide water.
Procedure. Prepare a saturated solution of hydrogen sul-
phide by bubbling the gas through distilled water. Place 50
ml. of standard 0 ·IN sodium arsenite into a 250 m!. volumetric
flask, add 20 ml. of the hydrogen sulphide water, mix well, and
add sufficient hydrochloric acid to render the solution dis-
tinctly acid. A yellow precipitate of arsenious sulphide is
formed, but the liquid itself is colourless. Make up to the
mark with distilled water, and shake thoroughly. Filter
the .. mixture through a dry filter paper into a dry vess'el.
Remove 100 m!. of the filtrate, neutralise it -with sodium
bicarbonate, and titrate against standard O·IN iodine and
starch to the first blue colour. The quantity of residual
e.rse1J.ious acid is thus determined, and is deducted from the
original 50 m!. employed.
Calculate the concentration bf the hydrogen sulphide in'the
'solution.
I M!. N As 20 a == ()'02556 g. H 2 S.
m, 120. Determination of ferricyanides.-Discussion. The
reactlOn between ferricyanides and potassium iodide is a
reversible one :
2K 3[Fe(CN).J + 2KI ~ 2K,[Fe(CN).J + 12,
436 Quan.titative Inorganic Analysis ,.
,
but proceeds quantita,tively from left to right in a slightly
acid medium in the presence of' a zinc salt. The very spar-
ingly soluble potassium zinc ferrocyanide is,formed :
2K 4 [Fe(CN),J + 3ZnS04 = K2Zn3[Fe(CN)sJ2 + 3K 2S04 ,
and the ferrocyanide ions are thus removed from the sphere
of action.
For practice, the student may determine the purity of
potassium ferricyanide, preferably of A.R. grade.
Procedure. ·Weigh out accurately about 10 g. of the potas-
sium ferricyanide and dissolve it in 250 ml. of water in a
volumetric flask. Pipette 25 ml. of this solution into a 250
m}. conical fl~sk, add about 20 ml. of 10 per cent potassium
iodide solution, 2 ml. of 2N sulphuric or hydrochloric acid, and
15 m!. of a solution containing 2·0 g. of crystallised zinc
sulphate. ,Titrate the liberated iodine immediately with
standard O'IN sodium thiosulphate and starch; add the
starch solution (2 m!.) after the colour has faded to a pale
yellow. The titration is complete when the blue colour has
just disappeared. When great accuracy is required, the
process should be conducted in an atmosphere . . of carbon
dioxide. Repeat the titation with two other 25 m!. portions
of the solution.
Calculate the purity of the salt.
I Ml. N Na 2 S 2 0 a == 0·3293 g. Ka[Fe(CN)6].
m, 121. Determination of ferric iron (iodometric method).

The reaction: •
2Fe+++
I
+ 2r ~ 2Fe++ + 12
is quantitative to the right hand side if the iodine is removed
by titration with sodium thiosulphate solution, but is nor-
mally too slow to be of practical utility. The addition of a
little cuprous iodide, however, acts as a catalyst and so
accelerates the reaction that titration may be carried 'oat
forthwith. The cuprous iodide is prepared as follows. About
1 m!. of N copper sulphate solution is treated with an excess
of 10 per cent potassium iodide solution (2 to 3 ml.) and the
free iodine is discharged by the addition of aqueous sodium
thiosulphate in the presence of starch solution. The precipi-
tate is first thoroughly washed by decantatlon, and then
shaken with 20 m!. of water; about I ml. of the suspension
thus obtained is used for a titration.
The slightly acid ferric solution, containing about 0·15 g. of
Fe, is treated with 1 ml. of the cuprous iodide suspension
and 10 per cent potassium iodide solution, and shaken well.
After 3 to 5 minutes the whole is titrated with standard O'IN
sodium thiosulphate to the disappearance of the blue starch
colour. The end point is fairly defini~e and permanent,
although the starch colouration reappears more rapidly than
in the copper-thiosulphate titration (Sections m, 105-106). ,
I ~l. N 'Na 2S20 a 7 0·05584 g. Fe~
• Note. If A.R..ferric alum is employed, the method will serve as a
useful check for the standardisation of thiosulphate solutions.
OXIDATIONS WITH POTASSIUM IODATE
m, 122. General discussion.""::'_A glance at the table of
standard oxidation potentials (Table XVI, Section I. 46)
reveals the fact that both potassium iodate and -potassium
bromate are stronger oxidising agents than iodine. The
,eaction Mtween potassium iodate and reducing agents, such
as potassium iodide or arsenious oxide, in weakly acidic
solutions, say 0·1 to 2'ON hydrochloric acid, stops at the
stage when the iodate is reduced to iodine:
KIO a + 5KI + 6HCl = 6KCl + 31 2 + 3H 20;
2KIO a + 5H3AsO~s + 2HCl = 2KCl + 5HsAsO, + 12 + H 2 0.
Actually the iodate is first reduced to iodide, and the latter is
subsequently converted into free iodi~e. ihis can readily be
shown by carrying out the oxidation of reducing substances
in weak acid' solutions (0·1 to 2N hydrochloric acid) containing
mercuric sulphate; here the iodide is converted into the fairly
stable complex ion [HgIJ- - as fast as it is formed, and
the r!?action is accordingly arrested at the iodide stage. An
example is:
4KIO a + 12HaAsOa + HgSO, (+ HCI)
=K 2SO; + K 2[HgIJ + 12HaAs~\ (+ HCI).
In mor y concentrated hydrochloric acid solutions (exceed-
ing 4N) L. W. Andrews (1903) has shown that the iodate is
reduced ultimately to iodine monochloride-the subst~nce
therefore acts as a very much more powerful oxidising agent.
Examples are* :
KIO a + 2KI + 6RCl = 3KCl + 3ICI + 3R20 ;
KIO a + 2RaA~Oa + 2RCI =21IaAsO, + KCI + ICI + RP ;
KIO a + 212 + 6RCI = KCI + 5ICI + 3H 20.
In "these reactions the iodine which is initially liberated is'
converted into iodine chloride. Now both iodine and iodine
chloride are brownish-yellow. in aqueous solution, but in
organic solvents, such as carbon tetrachloride or chloroform,
iodine is purple whilst rodine chloride remains yellow. The
point at which the last trace of iodine disappears may be
determined by the change in colour from purple to yellow
when a few m!. of carbon tetrachloride or chloroform are
shaken up in the solution during the titration (compare Sec-
tion 1.71 and Section m, 100). .
• These equations give the final products only.
438
The use of potassium iodate in the presence of high concen-
trations of hydroqhloric acid forms the basis of the Andrews
procedure for the determination of reducing agents. It has
been applied by G. S. Jamieson (1926)* and others to
numerous quantitative determinations. Essentially, all
methods depend upon the reaction:
r - 2€ ~ 1+,
and conditions are so chosen as to stabilise the iodine cation.
'In the Andrews procedure, the experimental conditions are
such that iodine 'chloride (ICI ~ 1+ + Crt) is stable and its
hydrolysis is repressed:
HCl + HIO -+- ICI + H 20 (or H+ + HIO --+ -l+ + H 20).
Let us write down the ionic equations for the various stages
in the reduction of iodate under the Andrews' conditions:
10 3- + 6H+ + 6.:' r + 3H 20, ,
or 1010 3- + 60H + + 60€ = lOr + 30H20;
'1:0 3 - + 6H+ + 51- = 31 2 + ;3H20,
or 210 3 - + 12H+ + lOr = 61 2 + 6H~0 ;
10 3- + 6H+ + 212 = 51+ + 3R 2 0,
or 31..0 3- + 18H+ + 61 2 = 151+ + 9H 20.
Upon addition, we have:
1510 3- + 90H+ + 60€ = 151+ + 45R2 0,
or 10 3- + 6H+ + 4€ = 1+ + 3Hp.
Hence under these particular conditions the equivalent weight
of potassium iodate is KIO a/4, or ~O'IN solution of potassium
iodate will contain KIOa/4 X 10, or 214'02/40, or 5·3505 g.
per litre.t • .
Other methods for stabilising the iodine cation have been
used and'give, in most instances, equally satisfactory results.
The most important of these is that due to R. La:ng (1922 --+)
in which cyanide ions are-used; in the latter case the acidity
should be slightly greater than N, and may be provided by
either hydrochloric or sulphuric acid. Here iodine cyanide
gives the necessary stability:
1+ + CN- ~ ICN .
... G. S. Jamieson. Vol.umetric Iodate Methods, 1926 (Chemical Catalog Co.,
N.Y.)
t The following reaction may occur: I
1+ + 2Cl- -= ICl.-.
t The same result can be deduced from the hypothetical equation:
KIO. + 2HCI = KCl + ICI + HIO + 20.
The simiiarity between the Andrews' hydrochloric acid method
and Lang's cyanide method may be illustrated by the stoich-
eiometric equations for the determinatio!l of iodides:
HI0 3 + 2HI + 3HCI = 3ICl + 3li 2 0;
HI0 3 • + 2HI + 3HCN = 3ICN + 3H 20.
In view of the highly poisonous character of hydrogen
cyanide, flasks with long and narrow necks must be used in
all processes employing the "iodine cyanide" method.
Only the "iodine chloride" procedure will, however, be
described in this book. *
Potassium iodate, as already pointed out in Section m,
lO2A, is available commercially in a high state of pUrity, is
anhydrous, can be dried at 120°C, and is stable in solution.
The A.R. produ,ct has an assay value of better than 99'9
per cent. The tit rations are carried out in glass-stoppered
bottles or flasks of 250 or 300 m!. capacity in the presence
of a'high concentration of hydrochloric acid, otherwise hydrol
lysis of the iodine monochlotide is liable to occur. The
solution to be titrated should contain so much hydrochloric
acid that, even after all the titrating·solution has been added,
the acidjty is at least 4N; that is, at least one-.third, and pre-
ferably one-half, of the solution should consist of concen-
trated acid. The end point is detected by the loss of the last
trace of purple (violet) colour from a layer'of 5 ml. of chloro-
form or carbon tetrachloride when shaken with the liquid.
When a titration' has actually been completed, there is no
return of the iodine colour "'even after keeping the solution for
a day. Iodate tit rations can be made in the presence of filter
paper, alcohol, forma1dehyde, saturated organic acids, as
well as many other kinds of organic matter. II
m~ lea! Preparation of O·lN (O'025M) potassium iodate.-

Dry some A.R. potassium iodate at 120°C for 1 hour and
allow it to cool in a desiccator. Weigh out exactly 5'35'1 g.
of the finely-powdered potassium iodate on a watch glass, and
transfer 'it by means of a clean camel hair brush directly into
a dry 1 litre volumetric flask,. Add about 400 to 500 m!. of
w~ter, and gently rotate the flask until the salt is completely
• For details of Lang's cyanide method, see A. D, Mitchell and A. M. Ward,

Modern Methods in Quantitative Chemical Analysis, Ijl3~, p, 6 (Longmans,
Green and Co.); also W. Bottger-R. E. Oesper, Newer Methods of Volumetric
Chemical Analysis, 1938, p, 76 (Van Nostrand and Co. or Chapman and Hall),
and F. Sutton-A, D. Mitchell, A Systematic Handbook of Volumetric Analysis,
1935 (J. and A. Churchill).
Volumetric Analysis -nl
diss9lved. Make up to the mark wjth distilled water. Shake
well. The solution will keep indefi!litely. ,
It must be emphasised again that the solution is O·lN only
for the reaction:
10 3 - + 6H+ + 4€ = 1+ + 3H 20.
ANALYSES .INVOLVING THE USE OF. POTASSIUM

IODATE SOLUTIONS.
m, 124. Determination of iodides.-This estimation IS
bas~d upon the reaction:
IO a- + 6H+ + 21- = 31+ + 3H2 0 .
(or, e.g., KIO a + 6HCI + 2KI = 3KCI 3ICI + 3H 20). -+
Thus 1 m!. 0'025M KIO s == 0·006347 g. r (since KIO a G
2r).
iodine in potassium iodide.
Weigh out accurately about 2·0 g. of potassium iodide, and
dissolve it in 250 ml. of water in a standard flask. Shake well.
Transfer 25' m!. of this solution by means of a pipette into a
stoppered reagent bottle of about 250 m!. capacity,· and add
25 m!. of water, jiO m!. of concentrated hydrochloric acid
together with about 5 m!. of carbon tetrachloride or chloro-
form. Cool to room temperature. Run in the standard
O·lN potassium iodate from a burette until the solution, which
at first is strongly coloured with iodine, becomes pale brown.
The bottle is then stoppered and vigorously shaken, and the
solvent layer acquires the purple colour due to iodine. Con-
tinue to add small volumes of the iodate solution, shaking
vigorously after:each addition, until the organic layer is only
very faintly violet. Continue the addition dropwise, with
shaking after each drop, until the solvent loses the last trace
of violet and has only a very pale yellow colour (due to iodine
chloride). The end point is very sharp and, after a little
experience, is rarely overshot .. If this should occur, a:·small
volume of the iodid~ solution is added from a graduated
pipette, and the end point re-determined. Repeat the
titration with two other 25 ml. portions of the iodide solution.
Calculate the percentage of iodine in the sample of potas-
sium iodide.
• Alternatively, a 250 m!. measuring flask with a short neck and a well-
fitting ground glass stopper may be used. The C9lour of the organic layer is
readily seen by inver!ing the flask so that the layer of indicator collects in the
neck.
Note. Iodine is similarly determined:
10 3- + 6R+ + 212 = 51+ + 2H;20,
or, e.g., KI0 3 + 6RCI + 212 = KCI + 5ICl + 3HzO.
.Hence 1 ml. 0'025M KIO s == 0·01279 g. I, since KIO., == 21 2 ,
A mixture of iodine and iodide is estimated by first titrating the
iodine with standard sodium thiosulphate solution, and then
determining the iodine + iodide in a fresh .portion"With potassium
iodate.
m. 125. Determination of arsenic or of antimollY.-Dis-
cussion. The determination of arsenic in arsenious com-
pounds is based upon the following reactions:
KI0 3 + AS 20 3 + 2HCl = KCl + AS 20 6 + ICI + H 20
(i.e., 103 - + AS 2 0 3 + 2H+ = 1+ + As S0 5 + H 20) ;
or KI0 3 + 2AsC1 3 + 5H 2 0 = ~Cl + 2H3As04 + ICl+4HCl.
Similar reactions.occur with antimonious cOny>ounds.
Thus 1 m!. 0'025M KIO a'==0'003746 g. As ===0'006088 g. Sb.
Iron and C'Opper do not interfere with these estimations.
AS 2 0 3 in commerci'3.l arsenious oxide or in the A.R. product.
Procedure.. Weigh out accurately about l·t g. of arseni-
ous oxide, dissolve this in a small quantity of warm 10 per
cent sodium hydroxide solution, and make up to 250 m!. in a
volumetric flask. Shake well. Transfer 25 m!. of this
solution to a 250 ml. stoppered reagent bottle, add 25 ml. of
concentrated hydro~hloric acid and 5 m!. of carbon tetra-
chloride. Titrate with the standard O·IN potassium iodate,
as detailed in Section m. 124. until the iodine colour in the
organic layer has disappeared. Allow to stand for 10 minutes,
and observe whether the carbon tetrachloride shows any
purple Colour. If no colour is seen, the titration is complete.
The a~idity of the entir~ mixture at the end of the titration
should not be less than aN and not mOf!! than 5N; with too
high acidity the reaction takes. place 'very slowly. Repeat
the titrat!on with two other 25 ml. portions of the solution.
Calculate the percentage of ASzO a in the sample.
m. 126. Determination of copper.--Discussion. The cop-
per is precipitated as cuprous thiocyanate, and then titrated
with standard potassium iodate. solution in the presence of
stron~ hydrochloric acid and a carbon tetrachloride or
chloroform indicator. The reaction proceeds according to
the following equq.tion :
7KIO a +2Cu 2(CNS)2 + 14HCI
= 7KCl + 4CuSO-, + 7ICl + 4!"fCN + 5H20.
Thus =
7KIO a .,= 4Cu, or I ml. 0'025M KIO a 0'009081 g. Cu.
The method may be applied to ores and certain alloys of
copper; the nitric acid employed for bringing the substance~
into solution must be expelled by evaporating with concen-
trated sulphuric acid to copious fuming.
copper in crystallised copper sulphate, preferably 'of A.R.
grade. About 0·02 g. of copper is a suitable quantity for
titration with 0·025M potassium iodate. .
Procedure. Weigh out accurately' about 0'8 g. of crystallised
copper ..sulphate CuS0 4 ,5H 20, dissolve it in water, add 5 ml.
of N sulphuric acid, and make up to 250 m!. in a volumetric
flask. Shake well. Transfer 25 mI. to a conical flask, add,
10 to 15 ml. of freshly-prepared saturated sulphurous acid
solution, heat to boiling and add 10 per cent ammonium
thiocyanate solution, slowly and with cohstant stirring, from
a burette until further addition produces no change in colour,
and add 3 to 4 m!. in excess (5 to 10 m!. in all). Allow the
precipitate to settle for 10 to 15 minutes, filter through a
quantitative filter paper or through a Gooch crucible with
asbestos. Wash with a cold I per cent ammonium sulphate
solution until free from thiocyanate. Transfer the washed
precipitate with the filter paper or with the asbestos quantita-
tively into a 250 ml. reagent bottle, add 30 mI. of concen-
trated hydrochloric acid, 20 ml. of water, and 5 ml. of chloro-
form or carbon tetrachloride. Add the O'IN potassium
io~ate rapidly whilst rotating the bottle in order to keep the
contents mixed. When the iodine has just disappeared from
the solution, insert the stopper and shake thoroughly. The
organic layer will be strongly coloured. From this point,
continue the titration slowly, shaking the closed bottle after
each addition of the iodate solution .until the iodine colour
has disappeal'ed from ~he carbon tetrachloride or chloroform
layer; this marks the end point. Repeat with two other
25 m!. portions of the solution.
Calculate the .percentag~ of copper in the sample used.
m, 127. Determination of mercury.-Discussion. The
~ mercury is precipitated as mercurous chloride. The latter
reacts with potassium iodate solution "Qnder the usual con-
ditions in accordance with the equation:
KI0 3 + 2Hg2 Cl 2 + 6HCl = KCl + 4HgCl 2 + ICI + 3H 20.
Thus KIO a == 4Hg == 2Hg 2Clz,
and 1 ml. 0'025M KI0 3 =0'02006 g. Hg=0'02361 g. IlgzCl 2 •
mercury in mercuric chloride, preferably of ~R~quality.
Procedure. Weigh out accurately about 2·5 g. of finely-
powdered mercuric chloride, and dissolve it in 100 m!. of
water in a measUfing flask. Shake well. Transfer 25 m!. of
the solution to a conical flask, add 25 ml. of water, 2 ml. of N
hydrochloric acid, and excess of 50 per cent phosphorous
acid solution. Stir thoroughly and allow to stand for 12
hours or more. Flter the precipitated mercurous chloride
through a quantitative filter paper or through a Gooch
crucible with asbestos, and wash the precipitate moderately
with cold water. Transfer the precipitate with Hie filter paper
or asbestos guantitatively to a 250 ml. reagent bott!e, add 30
ml. of concentrated hydrochloric acid, 20 m!. of water, and 5
ml. of carbon tetrachloride or chloroform. Titrate the
mixture with standard O'IN potassium iodate in the usual
manner· (Section m, 124) .
• Calculate the percentage of mercury in the sample.
m, 128. Determination of tin.-Discussion. Potassium
iodate solution reacts quantitatively with stannous chloride
in the presence of concentrated hydrochloric acid in accord-
ance with the equation:
KIO s + 2SnCl2 + 6HCI = E:CI + 2SnCl4 + ICI + 3H 20,
or I0 3 - + 2Sn++ + 6H+ = 2Sn++++ + 1+ + 3H zO.
Thus KI0 3 = 2Sn, or 1 rol. 0'025M KIO s = 0·005935 g. Sn.
This method possesses several advantages. Th~ end point
is very sharp. Extreme precautions need not be taken to
prevent atmospheric oxidation if the bulk of the iodate solu-
tion is added rapidly at first; the titration can then be com-
pleted at leisure. The method cannot be applied in the
presence ,of antimonYl cuprous and ferrous salts or precipi-
tated metals; these interfering substances can, however, be
simply removed.
tin in hydrated stannous chloride, preferably of A.R. grade.
Procedure. Weigh out accurately about 1·5 g. of crystallised
stannous chloride SnCl 2,2H 20, dissolve it in concentrated
hydrochloric acid, and make up to 250 ml. in a volumetric
flask. Shake thoroughly. Replove 25 ml. of the solution to
a 250 ml. reagent bottle, add 30 ml. of concentrated hydro-
chloric acid, 20 ml. of water, and 5 m!. of carbon tetrachloride
or chlon2form. Add standard o·IN potassium iodate rapidly
at first, whilst shaking the bottle to give the contents a gentle
rotary motion, until the iodine colour, which gradually appears,
has increased to the maximum intensity, th.en insert the
stopper, and sJ?ake the solution thoroughly. Continue the
titration slowly, shaking after each addition, until the organic
layer is no longer blue. Repeat the titration with two other
25 m!. portions of the solution.
Calculate the percentage of tin in the sample.
m, 129. Determination of peroxides (lead, barium, and
manganese dioxides).-Discussion. The method employed
is an indirect one. Excess of potassium iodide solution, which
has been standardised against potassium. iodate (Section m,
124), is added to a weighed amount of the dioxide, and the
mixture is titrated in the usual manner with standard potas-
sium iodate solution, using chloroform or carbon tetrachloride
as indicator. The reactions which occur are:
2M0 2 + 2KI + SHCl = 2MC12 + 2KCl + 2ICI + 4H 20
(where M = Pb, Mn or Ba)
i.e., 2M0 2 + 2r- + sH+ :_ 2M++ + 21+ + 4H 2 0 (oxidation) ;
KIO a + 2KI + 6HCI = 3RCl + '3ICI + aH 20,
i.e., I0 3- + 2J:- + 6H+ = al+ + 3H 20 (~ack titration).,
Now M0 2 :=; KI, so that when the excess of iodide is deter-
mined by titration with the iodate, the percentage of M0 2 in
the sample can be r~adily calculated. Any organic matter
which may be present does not interfere with the estimation.
Pb0 2 in the commercial dioxide.
Procedure. Weigh out accurately about 0·4 g. of lead dioxid~_
into a 250 mI. reagent bottle; it may be washed into the
bottle with a little water. Add 25 ml. of O·IN potassium
iodide- and a volume of qmcentrated hydrochloric acid· equal
to the volume of liquid already in the bottle. Insert the
stopper, and shake until all the lead. dioxide has dissolved.
Titrate the .excess of potassium iodide with standard O·J.N
potassium iodate as described in Section m, 124. Titrate-the
446 Quantitative Inorganic A nalys1's
potassium iodide solution against the standard potassium
iodate solution, and thence compute the amount of-potassium
iodide which has reacted with the lead dioxide.
C~lculate the percentage of Pb0 2 in the sample.
m,130. Determination of hydrogen peroxide.-Discussion.

The method depends upon the interaction of dilute hydrogen
peroxide solution with excess of standard sodium <arsenite
solution in the presence of sodium hydroxide:
..;\S2 0 3 + 2H 20 2 = As 2 0 6 + 2HaO.
The excess of sodium arsenite is then determined ,by titration

with potassium iodate as detailed in Section m, 125 :
KI0 3 + As 2 0 s + 2HCI = KCI + As 20 6 + ICI + H 20,
or IO s- + As 20 s + 2H+ = 1+ + AsaOs + H 20.
The amount of arsenite, found by titration, is dedu~ted from
that originally taken. Since As 2 0 s ;-:: 2H 2 0 2 , the percentage
of hydrogen peroxide in the original solution can be readily
calculated. This method, like that of Section m, 110, is not
influenced by the presence of organic preservatives.
Procedure. Dilute 25 ml. of "20 volume" hydrogen
peroxide to 250 m!. in a volumetric flask. Place 50 m!. of
0·1N sodium arse:p.ite (Section m,104A) and 10 ml.'of 10 per
cent sodium hydroxide solution into a 250 m!. reagent bottle~
and add 25 m!. of the dil1+ted hydrogen peroxide solution
slowly from a burette whilst the contents of the bottle are
gently rotated. Allow the mixture.to stand for 2 minutes,
and cautiously add 50 ml. of concentrated hydrochloric acid.
Insert the glass stopper and, whilst holding it down firmly.
shake the bottle vigorously. Release the stopper carefuliy
so as to allow the excess of pressure of carbon dioxide to
esca,pe without losing any of the .solution. Add 5 m!. of
shloroform or of carbon tetrachloride, and titrate the .unoxi-
dised arsenious acid with standard 0·025M potassium iodate.
Repeat the titration with two other 25 m!. portions of the
diluted hyqrogen peroxide solution.
Titrate 25 m!. of the arsenite solution against the standard
potassium iodate solution (Section m, 125). Calculate the
volume of the arsenite solution which has reacted with the
hydrogen peroxide, and thence the percentage of HaOa in
the.ori~nal solution.
1 Ml. N AsaOs == 0'01701 g. H 20 2 •

V olnmetric A na!ysis 447
m,131. Determination of hydrazine.-Discussion. Hydra-
zine reacts with potassium iodate under the usual Andrews
conditions thus:
KI0 3 + N2H, + 2HCI = KCl + N z + ICI + 3H 20,
or IO s- + NzH, + 2Ht = 1+ + N2 + 3H 20.
Thus KI0 3 '=:; N zH 4 ,
or ~ ml. 0'025M KIO s = 0·0008013 g. N2H4 = 0·003253 g.
N 2 H 4 , H 2S0 4 ,
As an exercise, the student may determine the N 2H4,H2SO,
content of hydrazine sulphate, preferably of A.R. quality.
procedure. Weigh out accurately 0·08 to 0·1 g. of hydra-
zine sulphate into a 250 m!. reagent bottle, add a mixture of
30 m!. of concentrated hydrochloric acid, 20 m!. of water and
5 ml. of chloroform or carbon tetrachloride. Run in the
standard 0'025M potassium. iodate slowly from a burette,
with shaking between the additions, until the organic layer
is just decolourised. .
Calculate the percentage of N 2H"H zS0 4 in the sample.
In, 132. Determination of thallium.-D2Scussion. Thallous
salts are oxidised by potassium iodate under the Andrews
conditions in accordance with the following equation:'
KIO s + 2TICI + 6HCl = KCl + 2TICla + ICI + 3H zO,
i.e., 10 3- + 2TI+ + 6H+ = 1+ + 2TI+++ + 3H20.
=
Thus KIO a 2Tl, or 1 ml. 0'025M KIO a == 0~01022 g. Tl.
Procedure. Place 20 ml. of the solution containing 0·25 to
0·3 g. of thallium (e.g., as thallous nitrate) into a 250 ml.
reagent bottle, add 60 mI. of concentrated 'hydrochloric acid
and 5 in]. of chloroform or carbon tetrachloride. Titrate with
standard 0·025M potassium iodate in the usual manner
(Section m, 124) -until the violet colour of the organic layer
just vanishes.
Calculate the thallium content of the solution.
m, 133. Determination of iodides by means of eerie sul-
phate.-Discussion. This estimation utilises an application
of .Andrews' principle (Section m, 122), the iodide being
oxidised to iodine chloride, ICI (~ 1+ + cq :
1- + 2Ce++++ = 1+ + .2Ce+++.
The concentration of hydrochloric acid at the end of the
titration should exceed 4N, i.e., two-fifths of the final volume
of liquid should consist of concentrated' acid.
448 Quantitative J norganic Analysis
the percentage of iodine in potassium iodide, preferably of
A.R. quality. . •
Procedure. Weigh out accurately about 2·0 g. of potassium
iodide, dissolve it in water, and make up to 250 ml. in a
standard flask. Place 25 ml. of this solution and 50 ml.
of concentrated l}ydrochloric acid in a 250 ml. glass-stoppered
reagent bottle containing 5 ml. of carbon tetrachloride.
Add standard O'IN ceric sulphate (Sections m. 74; 75)
from a burette with frequent shaking until tlie colour' of
the iodine is discharged from the solvent layer. Repeat the
titration with two other 25 ml. portions of the Iodide solu~ion.
C,!Jculate the percentage of iodine in the sample.
1 Ml. N Ce(S04)a == 0·06346 g. 1.
m, 134. Determination of vanadium with potassium iodate.
-Discussion. Vanadates are reduced by iodides in strongly
acid (hydrochloric) solution in an atmosphere of carbon
dioxide to the quadrivalent condition :..
2HaVO, + 2HI + 4HCl = 2VOCl a + 12 + 6H aO.
The liberated iodine and the,excess of iodide is determined by
titration with standard potassium iodate solution; the hydro-
chloric acid concentration must not be allowed -fo fall below
7N in order to prevent re-oxidation of the vanadium com-
pound by io~ine chloride.
HI0 3 + 212 + 5!lCl = 5ICl + 3H zO;
HI0 3 + 2HI + 3HCI = 3ICl + 3H zO.
The total result Qf the reaction is :
4H aV0 4 + 4HI + HIO a + I3HCl = 4VOCI z + 5ICl+15H zO,
and it can ,easily, be shown that I ml. 0'025M RI0 3 ==
0·005095 g. V.
This method is applicable in the presence of arsenate,
phosphate or ferric iron, and also in the presence of tungstic
acid, which may be held in solution by adding phosphDric
~~ ,
Procedure., 25 Ml. of the solution containing 0·05 to 0·10 g.
of vanadium (as vanadate) is placed in a 250 ml. glass-stoppered
reagent bottle, and a rapid current of carbon dioxide is passed
for 2 to 3 minutes'into the bottle, but not through the solution.
Sufficient concentrated hydrochloric acid is then added
through a funnel to make the solution 6 to 'SN during the
titration. A known volume (excess) of approximately O'05M
potassium iodide, which has been titrated against the stan-
dard iodate solution (Section IU, 124), is added. The whole is
mixed, allowed to stand for 1 to 2 minutes, 5 ml. of carbon
tetrachloride added, and then titrated as rapidly as possible
with standard O·025M potassium iodate u1).til no more iodine
colour can be detected in the organic layer. Concentrated
hydrochloric aCid should be added as needed during the
titration so that the concentraiion does not fall below 7N.
OlCIDATIONS WITH POTASSIUM BROMATE
ill, 135. General discussion.-Potassium bromate is a
powerful oxidising agent in acid solution (Tahle XVI, Section
I, 46). Reducing agents convert it smoothly into the
bromide: '
Br0 3- + 6H+ + 6€ = Br- + ~H20.
The equivalent weight is therefore 1/6 mol (KBr0 3 /6), or
167·'02/6, or 27'84, and a O·lN solution contains 2·784 g.
potassium bromate per litre. * At the end of the titration
free bromine appears:
Er0 3- + 5Er- + 6H+ = 3Er 2 + 3H 2 0.
The presence offree bromine, and consequently the end point,
can be detected by its yellow colour, but it is better to use
indicators such as methyl orange, methyl red or indigo-
carmine. These indicators have their usual colour in acid
solution, but are destroyed by the first exC,ess of bromine and
assume a pale yellow colour (yellow-green to yellow with a
reddish tinge for indigo carmine). The colour change is not
reversible, since, as already stated, the indicators are decom-
posed by the bromine. The quantity of bromate consumed by
the indicator is exceedingly small and can be neglected for
O'IN solutions. Direct titration with bromate solution in the
presence of irreversible dyestuff indicators are usually made in
hydrochloric acid solution, the concentration crf which should
be at least 1·5 to 2N. At the end of the titration some
chlorine may appear by virtue of the reaction:
10Cl- + 2Br0 3- + 12H+ = 5C12 + Br2 + 6H 20;
this immediately bleaches the indicator.
The titration.s should be carried out slowly so that the
indicator change, which is a. time reaction, may be readily
detected. If the determinations are to be executed rapidly,
the volume of the bromate solution to be used musf be known
approximately, since ordinarily with irreversible dyestuff
indicators there is no simple way of ascertaining when the
end point is close at hand. Examples of estimations utilising
direct titration with standard bromate are expressed by the
following equations:
• The same result can be obtained hy considering the hypothetical equation:
HBr0 3 = HBr + 30,
i.e., 1 mol of potassium bromate supplied G equivalents of oxygen.
450
Volitmetric Analysis 451
KBrO a + 3H aAsO a (+ HCI) = KBr + 3H aAs0 4 (+ HCI) ;
KBrOa + 3SbCl a + 6HCI = KBr + 3SbCIs + 3H 2 0 ;
KBrOa + 3SnCl 2 + OHCl = KBr + 3SnCl 4 :+- 3H 20 ;
2KBr0 3 + 3N~H4 (+ HCI) = 2KBr+3N z + OH 20( + Hel);
KBrO a + NH 20H (+ HCI) = KBr+HNOd-H20 (+ Hel);
KBr0 3 + 6K 4[Fe(CN)6J + OHCl
= KBr + 6Ka[Fe(CN)6J + 6KCl + :~H20.
Various substances cannot be oxidised directly with potas-
sium bromate, but react quantitatively with an excess of
bromine. Acid solutioI\s of bromine of exactly 'known con-
centration are readily obtainable from a standard potas.sium
bromate solution by adding acid and an excess of bromide:
BrO a- + 5Br- + 6H+ = 3Br 2 -I- 3H 2 0.
In this reaction one molecule of bromate yields six atoms of
bromine, hence the equivalent weight is KBr0 3 /6, identical
w,ith that of potassium bromate alone. Bromine is very
volatile, and hence such operations should be cd'uducted at as
Iowa temperature as possible and in reagent bottles fitteu
with ground glass stoppers. The excess of bromine is deter-
mined iodimetrically--by the addition of excess of potassium
iodide and titration of the liberated iodine with standard
thiosulphate solution:
2KI + Br2 = 2KBr + 12 ,
Potassium bromate is readily available in a high state of
purity; the A.R. product has an assay value of at least
99·9 per cent. The substance can be dried at 120° to 150°C,
is anhydrous, and the aqueous solution keeps indefinitely.
It can therefore be employed as a primary standard. Its
only c).isadvantage is that the equivalent weight is compara-
tively small.
~ 136. Preparation of O'lNpotassium bromate.;;;:-Dry som~
finely-powdered A.R. potassium brom~te for 1 to 2 hours at
120°C, and allow to cool ill a desiccator. Weigh out accur-
ately. 2·784 g. of the pure potassium bromate, and dissolve it
in 1 litre of water in a volumetric flask.

POTASSIUM BROMATE SOLUTIONS
m, 137. Determination of antimon~ or of arsenic,-Dis-
cussion. The antimony or .the arsenic must be present in the
452 Quantitative Inorganic' Analysis
trivalent condition. The reaction of trivalent arsenic or
antimony with potassium bromate may be written:
2KBrOa + 3M 20 a + 2HCl
= 2KCI + 3M 20 5 + 2HBr (M = As or Sb).
Hence I ml. N KBrOa :=; 0:04946 g. As 2 0 a :=; 0·03746 g. As
~ 0·07288 g. Sb 2 0 a :=; 0·06088 g. Sb.
The presence of tin and of considerable quantities of iron
and copper interfere with the estimations.
For practice in the use of potassium bromate solution, the
student may (a) titrate standard .. O·IN sodium arsenite
against the standard' bromate and (b) determine the percen-
tage of antimony in tartar emetic.
Procedure. Titration with standard sodium arsenite solution.
Place 25 m!. of standard O·IN sodium arsenite (Section m,
48,- Procedure A, Note I and Section m, 104A) into a conical"
flask, add 25 ml. of water, 15 m!. of concentrated hydrochloric
acid, and 2 dFops of methyl orange or methyl red indicator.
Titrate slowly with the standard O·IN potassium bromate with
constant swirling of the solution. As the end point approaches,
the bromate solution is added dropwise with intervals
of 2 to 3 seconds between the drops until the colour changes
sharply from red to colOUrless or very pale yellow. I Add
another drop of the indicator to make sure that the end
point has been reached. Repeat the titration with two other
25 ml. portions of the arsenite solution.
Compare the volume of potassium bromate solution used
with the calculated volume.
Procedure. AntimOny in tartar emetic. Weigh out accur-
ately about 4 g. of finely-powdered A.R. potassium antimonyl
tartrate or 5 g. of tartar emetic, dissolve it in water, and make
up to 250 ml. in a standard flask. Transfer 25 ml. of this
solution to a conical flask, add 25 m!. of water, 15 mI. of con-
centrated hydrochloric acid, and a few drops of methyl orange
or methyl red indicator. Add the standard O'IN potassium
bromate very slowly until the colour of the indicator is-just
discharged. Repeat the titration with two further 25 ml.
portions of the solution.
Calculate the percentage of antimony in the sample.
m, 138. Determination of metals by means of 8-hydroxy-
quinoline (" oxine ").-Discussion. Various metals (e.g.,
al~minium, iron, copper-, zinc, cadmium, nickel, cobalt,
Volumetric A nalysl:s 453
manganese, and magnesium) under specified conditions of pH
yield well-defined crystalline precipitates with 8-hydroxy-
quinoline. These precipitates have the general formula
Me(CgH60N)n' where n is the valency of the metal (see, how-
ever, Section I, 62L). Upon treatment of the oxinates with
dilute hydrochloric acid, the oxine is liberated. Oxine
reacts with 4 equivalents of bromine to give 5 : 7-dibromo-8-
bydroxyquinoline (R. Berg, 1926) :
C9 H 7 0N + 2Br2 = Cg H sONBr 2 + 2HBr.
Hence 1 mol of the oxinate of a divalent metal requires 8
equivalents of bromine, whilst that of a trivalent metal
requires 12 equivalents. The bromine is deri'yed by the
addition of standard O'IN potassium bromate and excess of
potassium bromide to the acid solution.
KBrOa + 5KBr + 6HCI = 6KCI + 3Br 2 + 3H 20.
Full experimental details are given below for the deter-
mination by the student of aluminium -in potash alum and of
magnesium in magnesium sulphate.
Of the other elements which' may be estimated volumetri-
cally with the aid of 8-hydroxyquinoline iron, copper (Sections
m, 106,. 126), cadmium (Section m, 38), zinc (Sections m,
145, 38), nickel ($ections m, 42, 38), cobalt (Section m, 37),
and manganese (Section m, 60) maybe determined by other
more convenient and less expensive methods. Outlines of
the procedures for cadmium, cobalt, nickel, titanium, zinc
and zirconium* will be given.
Procedure A. Determination of aluminium. Prepare a 2
per cent solution of A.R. 8-hydroxyquinoline in 2N acetic
acid (1) ; add ammonia solution until a permanent precipitate
is just produced, and re-dissolve this' by warming.
Weigh out accurately about 3·50 g. of potash alum
K2S04,AI2(S04)2,24H20, preferably of A.R. grade, and dis-
solve it in 250 m!. of water in a volumetric flask. Shake well.
Transfer 25 m!. of the solution, which contains about 0'02 g.
of AI, to a conical flask, add 125 m!. of water, and warm to
50° to 60°C. Then add a 20 per cent excess of the oxine
solution (1 m!. will precipitate 0·001 g. of AI), when the
complex AI(CgH60N)3 will be formed. Complete the preci-
pitation by the addition of a solution of 40 g. of ammonium
acetate in the minimum quantity of water, stir the mixture,
• For other elements, see R. Berg. Die Analytische Verwendung von 0-
Oxychinolin (" Oxin ") und Seiner Derivate, :!nd Edit.ion, 1038 (F. Enkc).
454 Quantitative I1w1'ganic Analysis
and allow to cool'. Filter the granular precipitate through a
coarse-mesh sintered glass crucib'le (or through a porous
porcelain crucible), and wash with hot water (2). DissolVe
the complex in warm concentrated hydrochloric acid,
or, better, in a IPixture of equal volumes of 10 to 15 per cent
hydrochloric aciu and alcohol, collect the solution in a 250
mi. reagent bottle, add a few drops of indicator (0·1 per
cent solution of the sodium salt of methyl red or 1 per cent
indigo-carmine solution), and 0·5 to 1 g. of pure potassium
bromide. Titrate slowly with standard 0'1N (i,e., M/60)
potassium bromate until the colour becomes pure yellow
(with either indicator). The exact end point is not easy
to detect, a8d it is better to add a slight excess of standard,
bromate solution (free bromine is thus present, and is
tested fqr by removing a drop of the liquid on to potassium
iodide-starch paper), diluting the solution considerably with
2N hydrochloric acid (to prevent the .precipitation of 5 : 7-
dibromo-8-hydroxyquinoline during the titration), then add
10 ml. of 10 per cent potassium iodide solution, and titrate the
liberated iodine against standard 0'1N sodium thiosulphate,
usjng starch as indicator (3). H.epeat the titration with two
other 25 m!. pbrtions of the aluminium- solution,
From the above discussion, it is evident that Al == 12Br.,
i.e., to 12,000 ml. of 0·1N bromate (or O·IN thiosulphate),
hence
1 Ml. N KBrO a 0:= 0'00~249 g. AI.
Calculate the percentage of aluminium in the sample
employed.
Notes. 1. An alcoholic solution cannot be used since the alumi-
nium complex is.slightly soluble in alcohol, nor can this solvent be
used for washing the complex. .
2. This will remove the excess of oxine. Complications due to
adsorption of iodine will thus be avoided.
3. A hrown additive compound of iodine with the dibromo com-
pound may scparatt' during the titration. It is, however, com!
pJetcly decomposed by'the thiosuJphatc, so that it does not affect
the end point. .
Procedure B. Determination of magnesium. Weigh out
accnrately about ]·0 g. of cryst!lllised magnesium sulphate,
prefcrahly of A.K 4uality, and llissolvl' it in :t;iO m!. of water
in a standard Ila.,k. l\lix thoroughly. Pipette 25 rn!. of
this solution into a beaker, add ;j g. of ammonium acetate
dissolved in 50 m!. of water, and heat to boiling. Add a slight
Volttmetrt'c A l1alysis 455
excess of a 2 per cent solutiu,Tl of oxine in 2N acetic acid, * and
then concentrated ammonia solution slowly and with stirring
until the liquid is faintly alkaline (e.g., to phenolphthalein
paper). Boil the mixture for 1 or 2 m~nutes until the yellow
precipitate becomes crystalline, and allow to settle.
The s'Qpernatant liquid should be yellow indicating that
an excess of reagent is present. Filter the precipitate
through a sintered glass crucible (lGa, or through a porous
porcelain crucible), and wash the precipitate thoroughly
with hot water (1). Dissolve the precipitate ·on the filter
in 2 to aN hydrochloric acid, and wash the filter with
the same acid until the washings are colourless. Transfer
the filtrate and washings quantitatively to a reagent bottle
or a glass-stoppered flask. Add 1 g. of pure potassium bro-
mide, several drops of 'indicator (methyl red or indigo-
, carmine), and titrate against standard O'lN potassium bro-
mate as described under aluminium (Procedure A). The same
difficulty in detecting the end point is experienced here also,
and is overcome in a similar way. Repeat the titration with
two other 25 ml. portions of the solution. .
Calculate the percentage of magnesium in the sample.
1 Ml. N KBrO a = 0·003040 'g. Mg.
Note. 1. The magnesium" oxinate" is appreciably soluble in hot water,
hente washing is discontinued when the colour of the washings changes from
a deep yellow to a pale yellow. It is probably better to employ a hot 1-2 per
cent ammoni,!- solu,tion as the wash liquid (R. Berg. 1938).
Procedure C. Determination of cadmium. The solution

(about 100, ml.; up to 0·1 g. Cd) of the metal is treated
with sodium carbonate solution until faintly cloudy, and
-the turbidity is lust removed by the addition of dilute acetic
acid. The mixture is warmed to 60°C, a to 5 g. of sodium
acetate are added (1), and a slight excess of a 2 per cent
solution of oxine in alcohol. stirred in, the supernatant liquor
being then of a golden-yellow colour and the precipitate of
Cd(C 9 H 6 0N)2,2H 20 a pale yellow. The mixture is heated
nearly to, boiling, set aside for a few minutes, and then filtered
on a sintered glass or porous porcelain crucible, washed with
hot water (2), dissolved in moderately concentrated hydro-
chloric acid (say, 4N), and titrated with O'IN potassium
bromate (Procedure A).
I Ml. N KBr0 3 ,~ 0·01405 g. Cd.
,
• FOf details of the preparation of the reagent, see Section It 620.
45G Quantitative I nOl'ganic Analysis
Note. 1. If copper is initially present, precipitation should be carried
out in 10 per ccnt acetic acid solution without the addition of
sodium acetate. The copper is thus completely precipitated as
Cu(C 9H 6 0N)2*, and the cadmium remains in the filtrate, from which
it is' precipitated by neutralisatiop. and the addition of sodium
acetate. Furthermore, precipitation in the presence of potassium
cyanide from sodium acetate solutions containing a trace 'of acetic
acid ensures that mercury, if present, remains in solution.
2. For the gravimetric detel'Jlliilation of cadmium, the precipitate
is dried at l30°C and weighed as Cd(C 9H 60N )2' which contains 28·05
per cen~ Cd.
Procedure D. Determination of' cobalt. The neutral or
weakly acidic (acetic acid) solutiont of the cobaltous salt
(150 ml. ; up to 0·06 g. Co) containing sodium acetate (5 g.) is
warpled to 70°C, treated with a slight excess of a 2 per cent
alcoholic solution of oxine, and boiled gently until the light-)
brown amorphous cobalt oxinate Co(C 9 H 6 0N)2,2H20 be-
comes crystalline. The latter is allowed to settle, filtered off
on a sintered glass or porous porcelain crucible, washed with
hot water, dissolved in fairly strong hydrochloric acid, and
titrated with standard O·IN potassium bromate (Procedure
A).
1 Ml. N KBrO a == 0·007368 g. Co.
Procedure E. Determination of nickel. The solution of 'the
nickelous salt (150 mI. ; up to 0·1 g Ni), containing 3 to 5 g.
of sodium-acetate and 8 to 10 per cent of acetic acid, is warmed
to 70°C, treated ,with excess of 2 per cent alcoholic oxine
solution, and then boiled until the precipitate of NJ(C 9 H 6 0N)2'
2H 20 becomes granular'. The whole is filtered through a
sintered glass or porous porcelain crucible, the precipitate.
thoroughly washed with hot water, dissolved in hydrochloric
acid, and titrated with O'IN potassium bromate (Procedure
A).
1 Ml. N KBr0 3 == 0·007335 g. Ni.
Note. This process affords a complete $eparation from manganese.
Procedttre F. Deterlnination of titanium. The solution
(150 ml.; up to 0·07 g. Ti) is treated with 1 g. of tartaric acid,
0·5 g. of sodium acetate, then with ammonia until barely
alkaline to phenolphthalein, and finally with 1·5 ml. of glacial
• For the gravimetric determination of copper, the precipitate· is dried at
105° to llO°C and weighed as Cu(CgH.ON). which contains 18·07 per cent Cu.
t If min(.'ral acid is pre~ent, soc.lium acetate is ac.ldec.l.
Volu1netric Analysis 457
acetic acid. The solution is warmed to 60°C, stirred whilst
excess of 2 per cent alcoholic oxine solution is added, and then
boiled for 10 minutes to effect granulation of the titanium
oxinate, TiO(C g H 6 0N}z,2HzO. The latter is then filtered off
on a sintered glass or porous porcelain crucible, and ma,shed
with hot water (I). The precipitate is dissolved in 1 : 1
hydrochloric acid, and titrated by O'IN potassium bromate
'and O'lN sodium thiosulphate (Procedure A).
1 Ml. N KBr0 3 ~ 0·005988 g. Ti.
Note. 1. The excess of oxine is removed when the washings are
colourless. To confirm this (or when in doubt) mix 25 m!. of the
filtrate with 10 m!. of concentrated hydrochloric acid, add 2 drops of a
a 0·2 per cent alcoholic solution of methyl red (or of the aqueous
solution of the sodium salt) and a little potassium bromide solution;
the addition of I drop of O·IN potassium bromate should result in
decolourisation. ..
Procedure G. Determination of zinc. The solution should
contain either (i) 5 g. of sodium or a,mmonium acetate and 4 g.
of acetic acid or (ii) 5 g. of sodium tartrate and 25 m!. of N
sodium hydroxide in 100 mI., and not-more than 0·15 g. of
zinc. In (i), the solution is warmed to 60°C, treated with a
slight excess of a 2 per cent solution of oxine in 2N acetic
acid or in alcohol, boiled for a few minutes, filtered and washed
with hot water (1). -In (ii) , the solution is treated with a 2
per cent alcoholic solution of oxine in the cold, warmed at
60° until the precipitate becomes crystalline, and filtered.
The well-washed precipitate (I) is dissolved in 2N hydro-
chloric acid and titrated with O'lN potassium bromate
(Procedure A).
1 Ml. N KBr0 3 =:; 0·008175 g. Zn.
Note. 1. For the gravimetric determination of zinc, the precipitate
is dried at 130° to ao°c and weighed as Zn(C UH&ON)2' which con-
tains 18·49 per cent Zn.
Procedure H. Determination of zirconi'QDl. Zirconium is
precipitated by oxine from an acetate-buffered solution of the
nitrate as Z.r(~H60N),: the-chloride and sulphate yield an
" oxinate " approximating to the composition ZrO(CuHsON)z.
'Fhe solution of zirconium nitrate in nitric acid should con-
tain not more than 0·015 g. of Zr and be free from' other
elements which react with oxine. The solution is evaporated
to a volume of a few m!. on the water bath, 10 ml. of nitric
acid (sp. gr. 1·20) added and again evaporated to a small
volume, but not to dryness. The solution is diluted to 150
Q*
458 • Qu,antitative Inorganic Analysis
mL, warmed to 50° to 60°C, 8 to 15 mL of a 2 per_cent alcoholic
soiution of oxine and 50 to 70 m!. of 2N ammonium acetate
solution added with constant stirring. The solution is boiled
until the precipitate is crystalline and pure yellow in colour.
The precipitate is allowed to settle and whilst still warm filtered
through a sintered glass crucible (lG4) or through a quantita-
tive filter paper, washed with hot 1 per cent ammonium
acetate solution until the washings are colourless, and then
with cold water (1). The" oxinate " is dissolved in cold 1 : 1
sulphuric·acid, the solution diluted to 150 m!. and titrated
with O'IN potassium broinate-(procedure A).
1 Ml. N KBrOa 0:= 0-005701 g. Zr.
Note. 1. For the gravimetric determination of zirconium, the
precipitate is dried at 130° to 140°C and weighed as Zr(C9H60N)4'
which contains 13·66 per cent Zr.
m, 139. Determination of hydroxylamine. The method
based upon the reduction of ferric solutions in the pre?ence of
sulphuric acid, boiling, and subsequent titration in the cold
with standard O'IN potas'!;ium perrnanganate frequently
yields high results unless the experimental conditions are
closely controlled:
2NH 20H + 2Fe z(SO,)a = 4FeS04 j- 2H zS0 4 + NzO + H 2 0.
1 Ml. N KMn0 4 == 0·01652 g. NH;0H.
The best results are obtained by oxidation with potassium
bromate in the presence of hydrochloric acid:
NHzOH -f HBrO a = HN0 3 + HBr + H 2 0.
The hydroxylamine solution is treated with a measured
volume of O·IN (1:.e., 111/60) potassium bromate so as to give
10 .to 30 ml. excess, followed by' 40 ml. of 5N hydrochloric
acid. After 15 minutes the excess of brumate is determined
by the addition of potassium iodide solution and titration
against standard O'lN sodium thiosulphate (compare Section
m,138).
1 M!. N KBr0 3 ;::= 0'005505 g. NH 20H.
OXIDATIONS WITH CHLORAMINE-T I
m, 140. General discussion. Preparation and standardisation
of O'IN solution.-Chloramine-T is the sodium derivativ.e of
N-chloro-p-toluenesulphonamide- CH a.C 6 H •. SO zNCINa,3H 2 0
(molecular weight, 281'70), and is obtainable commer-
cially in a high state of purity; it is a by-product in the
manufacture of saccharin. The aqueous solution reacts as if
it ,were a hypochlorite, and it has the great advantage that it
is far more
. stable than sodium hypochlorite solution: .
CHa.CeH•. SOzNClN-a + HzO ~ CH~,C6H4.S02NHz + NaOCI.
Chloramine-T ~as introduced as a volumetric reagent by A.
Noll (1924) as a cne?-p substitute for iodine in the determina-
tion of trivalent arsenic and antimony. The reagent reacts
in dilute acid (3 to 5 per cent) solution; if the acid is too
concentrated, secondary changes occur.
Chloramine-T reacts with potassium iodide in acid s8lution
to liberate iodine:
CH a.C 8 H•. S0 2NClNa + 2KI + 2HCl • /
= CH 3 .C 6 H•. S0 2 NH 2 + 1z + NaCI + 2KCl.

The equivalent weight is therefore 1)2 mol, or 281'70/2, or
14(11.85. Reaction also occurs quantitatively with arsenious
acid:
2CH 3 .C oH •. S0 2NClNa + ASzO a + 2H zO
= 2CH a.C 6 H,.SOzNH z + AszO. + 2NaCl.
•
Prepare an approximately Q·IN solution of chlor1mine-T
by dissolving 14·5 g. of the purest commercial product in 1
litre of wa'ter. I t may be standardised by one of two methods.
1. Treat 25 ml. of the chloramine-T solution with 2 ml. of
2N hydrochloric acid and 10 m!. of 10 per cent potassium
iodide solution. Titrate the liberated iodine with standard
o'IN sodium thiosulphate in Hie usual way, using 2 ml. of
starch as indicator.
2. Add the chloramine-T solution from a' burette to 25 m!.
of standard 0 ·IN arsenious oxide, containing a small crystal
of pure potassium iodide and 1 m!. of st"1.rch solution, until a
permanent blue colour just appear~.
469 .
ANALYSES INVOLVING THE USE OF .STANDARD
CHLORAMINE-T SOLUTIONS
mt 141. Determination of antimony.-For.practice in this
estim!ltion, the student may determine the percentage of
antimony in tartar emetic or in A.R. potassium antimony
tartrate.
Weigh out accurately about 0·35 g.. of tartar emetic or of
A.R. potassium antimonyl tartrate, and dissolve it in 100 m!.
of water in a conical flask. Add 0·2 g. of tartaric acid, 1 g.
of sodium bicarbonate, 2 mI. of starch solution, and a small
crystal of pure potassium iodide. Titrate with the stan-
dard 0-lN chloramine-T to the first permanent blue coloura-
tion. Repeat the titration with two similar quantities of the
antimony compound.
Calculate the percentage of antimony in the sample.
1 Ml. N chloramine-T ;:::; 0-0~088 g. Sb.
IDt 142. Determination of nitrites.-For practice in this
estimation, the student may determine the percentage purity
of sodium nitrite, preferably of A.R. grade.
Weigh out accurately about" 1·5 g. ot sodium p.itrite, and
dissolve it in 500 m!. of boiled-out water in a volumetric
flask. Shake thoroughly. Place 50 mI. of the standard
chloramine-T solution in a conical flask, and transfer 25 m!.
of the nitrite solution. to the flask by means of a pipette,
keeping the tip of the pipette below the surface of the liquid
during the addition. Add a few ml. of dilute acetic acid.
The sodium salt is decomposed with the separation .of the
corresponding free chloramine as a white cloud, which
rapidly diminishes in intensity. After 2 minutes, add ex.cess
of 10 per cent potassium iodide solution, and titrate the
liberateq. iodine with standard O'IN sodium thiosulphate,
using 2 m!. of starch solution as indicator. To another 50
m!. of the chloramine soluti9n add potassium iodide solution
and dilute acid, and titrate with the standard O"IN sodium
thiosulphate (Section W, 1(0); the difference between the
two tit res is a measure of the nitrite. Repeat the titration
with two other 25 m!. portions of the nitrite solution.
NaN0 2 + CH s·C.6 H 4 .s0 2NCINa + H 2 0
= NaNO a + CH a.C,H4 .s0 2NH 2 + NaCl.
Now CH a.C 6 H4 .s0 2 NCINa == 2Na 2S20 a == NaN0 2, hence
1 m!. N Na 2S2 0 s = 0·02301 g. N0 2 == 0·03456 g. NaN0 2 •
Calculate the percentage of NaNO~ in the sample.
Volumetric AnalJ:,.sis 461
m, 143. Determination of tin.-Discussion. The tin must
be present in the stannous condition. Reduction is best
effected in hydrochloric acid solution by means of pure
aluminium (free from iron), pure zinc turnings, or pure lead
foil. During the whole of the titration a continuous current
of carbon dioxide should be maintained above the stannous
solution, for this is very sensitive to atmospheric oxidation.
The reaction between stannous solution and the chloramine-T
is:
CH s.CII H 4 .S0 2NCINa + SnCl z + ·2HCl
= CH3,C6H,.SOzNH2 + Snel, + NaCl,
hence 1 ml. N chloramine-T = 0·05935 g. Sn.
tin in hydrated stannous chloride.
proced1~re. Weigh out accurately about 2·5 g. of crystallised
stannous chloride, preferably of A.R. quality, dissolve it in
150 mI. of concentrated hydrochloric acid, transfer to a 250
ml. volumetric flask, and make up to the mark with boi~ed-out
distilled water. Shake well. Transfer 25 ml. of the solution
(which contains about 0·15 g. Sn) into a 350 or 500 mI. conic'al
flask, add 30 ml. of concentrated hydrochloric acid and 100
mt of water, and finally pure lead foil to give a surface. of
about 24 square inches (or an equivalent quantity of A.R.lead
granules). Insert a three-holed rubber stopper as detailed
in Section m, 116, Procedure B, and pass a slow stream of car-
bon dioxide through the flask. Boil for 90 minutes and then
allow to cool; the current of gas is maintained throughout the
reduction* and the subsequent titration. Add sufficient
sodium hydroxide (14 to 15 g.), dissolved in air-free water, to
reduce the free acid concentration to 3 to 5 per cent in order to
avoid side reactions of the chloramine. .Then add a crystal of
potassium iodide, 2 ml. of starch solution, and titrate the mix-
ture with stand3;I'd O·IN chloramine-T until the presence of
the slightest excess is shown by the liberation of iodine and
the appearance of the usual blue colour. Repeat the estima-
tidn with another 25 ml. portion of the solution.
Note. If experience with a stannic salt is desired, stannic
ammonium chloride (ammonium chlorostannate) may be used.
... The reduction is not really essential in the present instance, but is des-
cribed in order to illustrate the technique involved in other estimations, e.g.,
tin in an alloy or in a stannic salt.
462 Quatttitath1c Inorganic Analyst's
ill, 144. Determination o! !elTocyanides.-Discussion.
Ferrocyanides are quantitatively oxidised by chloramine-T
in weakly aCid solution (2',i per cent by volume of hydro-
chloric acid) :
CH 3 .C 6 H 4 .S0 2NCIN·a + 2H+ + 2lFe(CN)6r---
= CHa.CaH4.S02NH2 + N<iCl + 2[Fe(CN)6r--.
Excess of standard chloramine-T solution is added to the
feebly acid solution of the ferrocyanide at 40°C; the residual
chloramine-T is determined by means of excess of standard
sodium' arsenite solution after the addition of sodium bi-
carbonate.
For practice, the student may determine the percentage
purity of A.R potassium ferrocyanide K 4 [Fe(CN)oJ, 3H 20.
Weigh Ot1t accurately about 1 g. of A.R. potassium ferro-
cyanide, and dissolve it in 100 ml. of distilled water in a glass-
stQPpered conical flask. Warm to 40°C, and add 2·5 ml. of-N
hydrochloric acid. Add 50 m1. of standard O'lN chloramine-
T and mix thoroughly. Then introduce 1 to 1·5 g. of sodium
bicarbonate, shake until dissolved, add a known volume
(exce~) of' standard O·lN sodium arsenite. Determine the
excess of the latter by the addition of a crystal of potassium -
iodide, and 2 ml. of -starch solution, followed by titration
with the standard chloramine-T solution to the first per-
m:tnent blRe colour.' Repeat with a similar quantity of
potassi.um ferrocyanide. ..
Calculate the percentage purity of the sample.
1 M1. N chloramine-T -= 0·3684 g. K4[Fe(CN)6l
MISCELLANEOUS VOLUMETRIC DETERMINATIONS
m, 145. Determination of zinc with standard potassium
ferrocyanide solution.-Discussion. Zinc ions in neutral or
acid medium react with potassium ferrocyanide solution to
form the very sparingly soluble potassium zinc ferro cyanide :
3Zn++ -+- 2K 4 [Fe(CN)6J = K2Zna[Fe(CN)6J2 + 6K+,
Formerly the end point of the reaction was determined by
means of external indicators, such as uranyl nitrate and
ammonium molybdate. The acidified zinc solution was
slowly titrated at 60° to 80°C with the ferro cyanide solution
until a drop of the solution, brought into contact with a drop
of uranyl nitrate solution on a white .spot plate, gave a light
brown colour (due to the formation of uranyl ferrocyanide).
In recent years, the external indicator has been replaced by the
more convenient internal indicator (diphenylamine, diphenyl-
benzidine, or sodium diphenylamine sulphonate). The latter
substances are .oxidation-reduction indicators, and are there-
fOIe dependent for their action upon the ratio of the concen-
tration of ferricya:qide to ferro cyanide in the solution (compare
Section 1,50). A 0·05M potassium ferro cyanide solution, to
which 0·3 g. of potassium ferricyanide per litre is added, is
employed; the solution is accordingly approximately
O'OOIM with respect to ferricyanide. The oxidation potential
of a ferrocyanide-ferricyanide electrode is given by the equa-
tion (Section I, 48) :
° [{Fe(CN)6} ---] 0
E = E + 0·0591 log [{Fe(CN)6} ] (at 25 C).
As long as excess of zinc ions remain in solution, the concen-
t,ration of the ferro cyanide is very small, and the potential is
large. As soon as the zinc ions are quantitatively precipi-
tated, however, the next drop of ferrocyanide solution causes
a sudden increa:se in [{F~(CN)6r - - -], and hence a sudden
decrease in the oxidatLon potential. The .end point may
therefore be detected by any of the three indicators men-
tioned above. We may also titrate in the Ieverse manner;
the method can therefore be used for the titration of ferro-
cyanides by means of standard zinc solution.
In this estimation, it is essential to work, so far as possible,
under uniform conditions, if concordant results are to be
obta,.ined. The titration must not be carried out too rapidly,
46:}
464 Quantitative Inorganic A.nalysis
and the solution must be thoroughly shaken throughout the
titration, otherwise over-titration may ea:sily occur. Nitrates,
oxidising agents, lead, copper, cadmium, iron, manganese,
nickel and cobalt must be absent.
Procedure. Preparation of O'05M POtassium ferroc:vanide.
Weigh out 21·12 g. of A.R. potassium ferrocyanide
K,[Fe(CN)8]' 3H 2 0 and 0·3 g. of A.R. potassium ferricyanide,
dissolve them in water, and make up to I litre in a volumetric
flask:
If the potassium ferrocyanide is pure,
1 ml. 0:..Q5M K,[Fe(CN)6] =
0·004904 g. Zn.
Standardisation of the potassium ferrocyanide solution. Prepare
a O'IM zinc solution by weighing out either 1·6345 g.
of A.R. zinc or 2·0345 g. of ignited A.R. zinc oxide, dissolving
either in diluteJ>ulphuric acid, and making up to 250 ml. in a
volumetric flask. If exact weights are not employed, the
exact concentrations may be calculated from the weights used.
Pipette 25 ml. of the zinc solution into a 250 to 350 ml.
conical flask. add 50 m!. of water, 25 m!. of 7N sulp4uric acid.
10 g. of ammonium sulphate, and several drops of diphenyl-
benzidine solution (Section In, 63) *. Titrate the cold soluti9b
slowly and with vigorous shaking until the colour change from
blue-violet to pale green is permanent. Repeat the, titration
with two other 25 m!. portions of the zinc solution.
Calculate the volume of the potassium ferrocyanide solu-
tioll: equivalent to 0·01 g. of zinc, and compare this with the
value deduced from the actual weight of A.R. potassium
ferrocyanide used.
For further practice in this estimation, the student may
determine the percentage of zinc in crystallised zinc sulphate •.
preferably of A.R. quality, or of zinc in commercial zinc
oxide. The sample must be treated exactly as in the stan-
dardisation.
m t 146. Determination of pbospborus.-Discussion:- The
phosphorus must be in the form of .orthophosphate. Upon
treatment at 200- to 45°C with a large excess of ammonium
molybdate solution in the presence of nitric acid, ammonium
phosphomolybdate (NH4laP04,12MoOa,2HNOa,H20 is pre-
cipitated; 'this is converted into (NH4)3P04' 12MoO a
when suitably washed with dilute potassium nitrate solution.
* If diphenylamine is used, 3 drops of the indicator are added; with more
of the indicator, the results are low. The amount appears to be immaterial
(up to 10 drops) for diphcnylbcnzi<line.
Volumetrt'c A ';talysis 465
The phosphorus in the washed precipitate may be deter-
mined in one of two ways.
A. It may be titrated against standard sodium hydroxide
solution with phenolphthalein as indicator. The following
reaction is assumed to take place·:
(NH4)3PO,,l2Mo03 + 23NaOH
= llNa 2MoO, + (NH4)2Mo04 + Na(N)I,)HP0 4 + llH 20,
whence 1 ml. N NaOH == 0·001349 g. P == 0·003088 g. P 2 0 6 •
B. The precipitate is dissolved in dilute ammonia solution,
sulphuric acid added, and the" resultant solution passed
through a Jones reduct or (the molybdenum is thus reduced
to the tervalent state) into excess of ferric alum solution.
The ferrous salt is titrated with standard O·lN potassium
permanganateo
1 Ml. N KMnO, = 0·0008617 g. P,
since P == 12MoO a, which is equivalent to 18 atoms or 36
equivalents of oxygen.
Procedure. Prepare the molybdate reagent as follows. Dis-
solve 100 g. of A.R. molybdic anhydride or 118 g. of A.R.
molybdic acid (ca. 85 per cent Mo0 3 ) in a mixture of 400 ml.
of water and 80 ml. of concentrated ammonia solution; filter,
if necessary. Add the molybdate solution thus obtained
slowly and with constant stirring into a solution containing
400 mi. of concentrated nitric acid and 600 ml. of water; the
end of the tube carrying the molybdate solution should dip
under the surface of the dilute nitric acid. Keep the mixture
in a warm place for several days, or until a portion heated to
40"" to 45°C deposits no yellow precipitate. Decant the
solution from any sediment, and preserve in glass-stoppered
bcltl~ •
• For practice in this estimation, the student may determine
the percentage 'of phosphorus in anhydrous sodium phosphate
Na 2 HPO" preferably of A.R. grade: this should be dried for
I hour at 120°C and allowed to cool in a desiccator before use.
Weigh out accurately 0·2 to 0·25 g. of anhydrous sodium
phosphate, dissolve it in water, and dilute to 250 ml. in a
volumetric flask. Shake well. Transfer 2iJ ml. of the solu-
tion into a 250 to 350 mr. conical flask, dilute to 100 mI., and
add 12 g:. of pure ammonium nitrate; shake until the solid
dissolves. Add 75 ml. of the molybdate reagent, previously
warmed to 40° to 45°C, slowly using a tap funnel or a drawn-
out funnel, and shaking continuously. It is important not
466 .Quantitative 11torganir A nalY~is
to heat above 45°C in order to avoid contamination of the
precipitate with' molybdic anhydride. Fit the flask with a
rubber stopper, shake vigorously for 10 minutes, and allow to
stand for 30 minutes. Filter the precipitate through a
quantitative filter paper. Wash the flask and the precipitate
with I per cent potassium nitrate solution until the filtrate
gives no acid reaction with methyl orange; this will require
about 100 mi. of the washing liquid. Place the paper and
precipitate in the original flask. Complete the determination
by either of the following methods.
A·. Treat the precipitate (NH4)aP04,12MoOa with a slight
excess of O'IN carbonate-free sodium hydroxide (say, 50 mI.).
Stopper the flask and shake; if all the precipitate does not
dissolve, a further quantity of standard alkali solution must
be added. Dilute to approximately 150 mI., add 5 drops of
phenolphthalein indicator, and titrate with standard O·IN
hydrochloric acid until the pink colour is completely dis-
charged. Complete the titration by adding standard alkali
until the re-appearance of the pink colour.
Calculate the volume of standard alkali which has reacted
with the precipitate, and thence the percentage of phosphorus
in the original salt.
B. Dissolve the precipitate" in 2N ammonia solution, and
add excess of. sulphuric acid until the concentration of the
latter is about 2N. Pass the solution at once through a' ] ones
reduct or (Secti9n m, 51B) into 50 m!. of A.R. ferric alum
soluti9n (100 g. of A.R. erysta.ls, and 25 ml. each of concen-
trated sulphuric acid and syrupy phosphoric acid in 1 litre).
Wash with about 200 m!. of water, and titrate the ferrous salt
with standard O'lN potassium permanganate. Carry ouf a
.
blank on the acid, water, and ferric alum solution .
Calculate the percentage of phosphorus in the original solid.
In, 147. Determination of temporary and permanent hard-
ness of water.-Discussion. Waters are describe<;l. as soft"
It
or hard" according to their action upon soap. A water is

It
said to be " soft" if it gives an immediate lather with soap

solution, and. " hard" if a lather is obtained with difficulty.
Hardness is due ahnost entirely to the presence of calcium and
magnesium ions; these combine with the anion~ of ,the soap
(usually sodium or potassium salts of oleic, palmitic or stearic
acids) to form ins""oluble calcium and magnesium compounds.
The actual salts present in hard water are the bicarbonates,
chlorides, and sulphates of calcium and mag-nesium.
Upon boiling hard water, the bicarbonates are decomposed
with the precipitation of the normal carbonates:
Ca(HC0 3 )2 = CaeO a + H 2 0 + CO 2 • ...
The hardness due to the presence of alkaline-earth bicarbon-
ates therefore largely disappears on boiling (a little of the
resultant oarbonate dissolves in the water); this hardness -is
designated as temporary hardness. The residual hardness,
due to the chlorides and sulphates of the alkaline earths and
the amount of their carbonates soluble in pure water, cannot
be removed by boiling and is termed Permanent hardness.
The sum of the temporary and permanent hardness of a water
represents the total hardness.
The most accurate method for the estimation of the
total hardness is to determine the calcium and magnesium by
a gravimetric procedure; the result is expressed in terms of
calcium carbonate. Two other procedures, Clark's and the
titration method, are in general use, but these do not give
identical results. Clark's method is -the older process;
accurate results are not obtained if a large quantity of mag-
nesium salts is present, but this is rarely the case in actual
practice. It has the advantage. however, that it does measure
the soap-consuming power of the water. The titration
method gives better results on the whole, since magnesium
salts are quantitatively precipitated when it is used. •
Hardness is conveniently expressed as parts of CaCO a per
100,000 parts of water; if a solution contains n parts of CaCO a
in 100,000 parts of water, it possesses n degrees of hardness.
This is the standard employed in France, and will he adopted
in this book. In Germany, each gegree of hardness indicates
I part of CaO in 100,000 parts of water. The term" English
degree of hardness" is sometimes applied to the hardness
based upon grains of CaCO a per gallon of water (1 Imperial
gallon = 70,000 grains). In the United States, hardness is
given in terms· of grains of CaC0 3 per U.S. gaUoh, which is
five-sixths as large as the Imperial gallon·. ~ Henee: 1
" French" degree = (Hio .. German ",degree ~-,~ 0·70 " En-
glish" degree = 0·li8!) .. U.S." degree. When magnesium
salts are present, they are expressed in terms of the equivalent
amounts of CaC0 3 or of CaO.
Procedure A. Clark's method. This method must be
carried out with care and slowly, especially for waters con-
taining magnesium salts; the eXperimental conditions must
be closely followed., Before commencing the determination,
468 Quantitative I norga1tic Analysis
it is advisable to estimate the total solid matter present by
evaporating a known volume of the water, and weighing the
residue. This gives a useful approximate idea as to how much
to use for the soap test. The total hardness for many waters
is about one half the total solids, and thesarnple may be so
qiluted, if necessary, as to bring it within the limits men-
tioned below. Thus if a water contains 50 grains of solids per
gallon (i.e., ca. 70 parts per 100,000), take 25 m!. for hardness.
The tedious and time-consuming evaporation of a known
volume of water can be avoided by employing the volumetric
trial and error method. Thus determine whether 50 'ml. of.
the water will produce a lather with 15 mI. of standard soap
solution (details are given below); if no lather is produced,
try 20 ml. of the soap solution, and continue the process until
the a_pproximate volume of soap solution required to produce
a permanent lather is ascertained.
Preparation of standard solution of calcium chloride. Weigh
out accurately 0·2000 g. of pure oalcite (or of A.R.
calcium carbonate) and dissolve it in dilJ.lte hydrochloric acid,
taking care to keep the vessel covered with a clock glas~ to
prevent loss by spirting. Use a platinum dish if available,
otherwise eIJlploy a Pyrex vessel. When all the solid has
dissolved, evaporate to dryness on the water bath, add a little
.distilled water, and again evaporate to dryness. Repeat the
evaporation several times to ensure complete expulsion of the
free acid. Finally, dissolve the residue of calcium chloride
in distilled water, and make up to 1 litre in a standard flask. .
50 Ml. of this solution correspond to 10 mg. of CaC0 3 •
Preparation of standard solution of soap. Method 1. Weigh
out 10 g. of soft soap (Sapo moUes, B.P.)*, and dissolve it in 1
litre of a mixture of equal volumes of industrial methylated
spirit and distilled water.
Method .2. Weigh out 50 g. of oleic acid into a beaker, and
add 100 mI., of alcoholic potash, made by dissolving 20 g. of
potassium hydroxide (sticks or pellets) in 180 ml. of industrial
spirit; continue adding the same solution from a burette
until a drop of the oleate just gives a red colour with phenol-
phthalein spotted on a white plate-about 10 ml. more being
required. Make the volume up to 500 ml. by the addition of
-the methylated spirit. 56·25 Ml. of the solution thus ob-
tained are diluted with a mixture of industrial methylated
spirit (2 vols.) and water (1 vol.) fo 1 litre.
* For elementary work. the so·called " Castile" soap is satisfactory.
Volmnetric Analysis 469
In either case, the solution is set aside in a cool place
(preferably in an ice chest) for 24 hours, filtered through a
double filter, and standardised against the standard calcium
chloride solution as described below. The soap solutions will
be a little too concentrateQ., and are to be diluted to such
strength that exactly 14·25 ml. are required to form a per-
manent lather with 50 ml. of the calcium chloride solution.
Standardisation of the soap solution. Measure out 50 ml. of
the standard calcium chloride solution into a glass-stoppered
bottle of about 250 ml: capacity. Run in 1 m!. of the soap
solution from a burette into the bottle, close the bottle, and
shake it vigorously for a short time. If no permanent lather
is formed, add another 1 mI., and shake again. Continue the
addition of the soap solution, shaking the mixture well after
each addition, until a lather is produ..ced which remains for a
short time when the bottle is laid upon the bench. The,
titration is complete, when the lather remains upo.n the sur-
face in an unbroken layer for 5 minutes. even when the bottle
is rolled half-way round on its side. To~ards the end of the
titration, the volume of sqap solution which is added each
time should be decreased, and it should finally not exceed
0·2 to 0·3 m!. Repeat the titration with another 50 m1. of
the standard calcium chloride solution; add the soap solution
1 ml. (or less) at a time, with shaking after each addition, and
never in large quantities.
Dilute the soap solution with the calculated volume of the
water-methylated spirit mixture such that exactly 14·25 m1.
of the final solution are required to produce a lather with 50
ml. of the standard calcium chloride solution.
Determination of the total hardness. Measure out 50 ml. of
the water, or, if necessary, a smaller quantity* (usually 25 ~l.
or 10 mI., together with 25 ml. or 40 mI. of recently boiled
and cooled distilled water) into a 250 ml. glass-stoppered
bottle. Titrate the water with the soap solution as described
under the standardisation. Waters containing much mag-
nesium salts give a false or " ghost" lather after the addition
of only a few mI. of soap solution; this, however, disappears en-
tirely on allowing the bottle to remain for several minutes on
its side after an extra vigorous shaking. Such water must always
be diluted so that not more than 7 m!. of -$oap solution are
required to produce a permanent lather. Repeat the titration.
• A titre of less than 16 mI. of the soap solution should be obtained. The
requisite volume of the water is obtained either from an estimation of the
total solids or by a preliminary titration with 50 ml.. of the water.
470 Quant£tative Inorganic Analysis
Ascertain the hardness from Table XXIV;. multiply the- -
result by 2 or 5 when 25 ml. or 10 ml. of the water, diluted to
50 ml., have been used. It should be pointed out that some
authors prefer to regard th~ actual titre as the real measure
of the hardness, and consider t~c Table of little significance
for finding the true amount of CaC0 3 •
TARI.E XXIV. TARI.E OF HARDNESS. PARTS IN 100,000.
(Calculated fortitrations of ;'i0 ml. of water.)
MI. of CaCO.
soap per
I MI. of
soap
cacoall MI.
per soap
CaCO.
per
o~ MI. of CaCO.
soap per
solHtion ](/0,000 I solution 10U,O(/(/ solution 1UO.000 solution 100.000
----- ----
- 4·,;1
-----
!l'0
- - H) 10·30 12·0 10·43
-- - 4·1 4·7L !H HH5 12·1 lli·39
- - 4·2 4·!l(l 8-2 10·60 ]').,) 16'75
- '- 4'3 5·00 8·3 10·75 12·3 16·90
- - 4·4- 1>-14 8·4 ]().!l\) 12·4 17·06
- --- 4·5 5·29 8'5 1l'05 12·5 17·22
--
U·i
-
0·00
4·(; •
4'i'
5·43
5·57
8·6
8'7
11·20
11·35
12·6
12·7
17·38
17·54
0·8 0·16 4·8 5·71 g·8 ll·5O 12·8 17·70
0·0 0'3:! 4·9 5·86 8'9 11·65 12'9 17·86
1'0 0·48 5'0 6·00 9·0 n'80 13·0 18·02
1·1 0'63 lH 6·14 !)ol 11·9j 13·1 18·17
1·2 0'79 5·2 6·29 9'2 12·11 13·2 18·33
I·;) 0·95 1i·3 6·43 0·3 12·26 1f '3 18'49
1'4 HI 5·4 6·57 9'4 12·41 13·.j. 18·65
Hi 1'27 0·5 6·71 9'5 12·56 13·5 18'81
1'6 1·43 0·6 6·86 9'6 12·71 1:1'6 18·97
1·7 1'56 5·7 7'00 9'7 12·86 13·7 19·13
1·8 1·69 5·8 7·14 9··8 13·01 13·8 19·29
1·9 1-82 0·9 7·29 9·9 13·16 13·9 Ig·44
2·0 1·95 6·0 7·43 10·0 13·31 14·0 11}-60
2·1 2·08 fH 7·57 10·1 13·46 14·1 19-76
2·2 2·21 6·2 7·71 10·2 13·61 14·2 19·92
.2·3 2·34 6·3 7·86 - 10·3 la'76 14·3 20:08
2'4 :!·47 • 6·4 g·OO 10·4 13-91 14'4 20·24
2·0 2·60 6·5 !H4 10·5 14·06 14·5 20·40
!H) 2'73 6·6 8·29 10-6 14·21 14'6 20·56
2·7 2·86 6-7 9'43 10'7 14·37 14'7 20·71
2·8 2·99 6·9 8·57 10·8 14'02 14·8 20·87
:!·9 3'12 6·!) 8'il 10·9 14·68 14'!) 21·03
3·0 ~-25 7·0 8·86 11·0 14·84 15·0 21·19
3·1 3·38 7·1 9·00 11-1 15·00 15·1 21·35
3·2 3'51 7·2 9·14 11·2 15·16 15·2 21·51
3-3 3·64 7·3 9·29 1.1·3 15·32 15·3 21·68
3·4 3'77 7·4 9·43 1l·4 15·48 15·4 21·85
~·5 3·90 7'0 1)'07 11·5 15'63. 15·5 22·02
3·6 4·0:1 j·6 9·71 IHi 15·79 15'6 22·18
4·16 7·7 9·86 11·7 15·95 15'7 22·35
3'7
3·8 4·29
- 7·8 10·00 11·8 16·11 15,8- 22·52
:l·9 4·43 7·9 10·15 11·9 I 16·27 15·9 22·69
- - - , -
I
- I
I
!
- 16·0 22·86
Deterniination of permanent hardness. Boil a known volume
(say, 100 m!.) of the water gently in an open flask for
30 minutes. Mark' the original revel of the water in the
flask, and add hot distilled water from time to time to make
up the loss by evaporation. At the end ·of 30 minutes, cover
the flask loosely and cool to the laboratory temperature, then
make up to the original volume by the addition of recently
boiled and cooled distilled water. Filter through a dry filter,
and determine the hardness in the filtrate by titration with
standard soap solution.
Determination 01 temporary hardness. The difference between
the total hardness and the permanent hardness gives the
temporary hardness of the water.
Procedure B. Titration method. Place 500 m!. of the
water, or a smaller volume (2150 or 100 m!.) if the water is very
hard, into a large evaporating dish or casserole, add a few
drops of methyl orange or, pr:eferably, methyl orange-indigo
ca.rmine indicator, and add standard O'IN hydrochloric acid
from a burette until the end point is reached. It is advisable
to add the same amount of indicator to 500 m!. of distilled
water and introdUce the st..andard acid until the "acid"
colour IS produced.' The tint can then be matched by that
of the water under examination; the blank correction is also
determined at the same time: Subtract the blank correction
from the volume of standard acid used, and thence calculate
the corresponding weight of caro s and convert it into parts
per 100,000.
Ca(HC0 3)2 + 2HCI = CaGl 2 + 2H 2 0 + 2C0 2,
t.e., I ml. O·lN HCl == 0'00.5005 g. CaCo a•
Z;ote. The number of mI. of acid used is often regarded as the
temporary hardness. This, however, includes some calcium car-
ponate which would remain in solution after boiling, and it is better
to term it the alkalinity. An empirical deduction of 0·4 ml. for
o·IN acid is then made to allow for the solubility of the calcium
carbonate, a,nd the differen:e represents temporary hardness.
Determination of permanent hardness. Measure out 250
m!. of the water into a conical flask, best of Pyrex gla,ss, and
boil for 30 minutes. Add an excess (say, 50 ml.) of standard
O·lN sodium carbonate, and evaporate to dl1yness on the
water bath. Take up the residue in a little boiled-out dis-
tilled water, filter through a dry quantitative filter paper, and
wash the residue four times with hot water. Allow the filtrate
to cool, and make it up to 250 m!. in a volumetric flask with
472 Quantitative Inorganic A natysis
boiled-out distilled water. Titrate 50 ml. or 100 ml. with
standard O'IN hydrochloric acid, using methyl orange or
better, methyl orange-indigo carmine indicato,r.
Calculate the volume 'of residt1al O'IN sodium carbonate,
and from this the volume of (J·IN sodium carbonate consumeo.
in precipitating the alkaline earth salts of the strong acids.
CaSO, + Na 2CO a = CaCO a + NazSO,;
MgCl z + Na 2CO a = MgC0 3 + 2NaCl.
1 Ml. O·IN NazCO a ;:::= 0'005005 g. CaCO a.
Calculate the permanent hardness of the water (parts of
CaCO a per 100,000).
m •. 14S. Determination of sodium.-The sodil1m is precipi-
tated as NaZn(U02)a(C2HaOz}g,6HzO or as NaMg(UO Z}3
(C ZH 3 0 2}9' 6'5HP (Section IV, 47B and O), and the uranium
determined according to Section ~; 91.
m, 149. Determination of potassium.-The potassium is
precipitated as K 2Na[Co(N0 2}6] (Section IV; 480), and tne
nitrite in the precipitate determined with cerie sulphate
(Section m, 79).
m, 150. Determination of cadmium.-the cadmium is pre-
cipitated as the .B-naphthaquinoline complex [(C13HpN)2H2]CdI,
(Section IV, ISB). The precipitate is decomposed with
ca. 20 ml. of 2N ammonium or sodium hydroxide, and the
solution treated successively with ~O ml. of .2N hydrochloric
or sulphuric acid, 5 ml. of 10 per cent potassium cyanide
solution, and 2 ml. of starch solution. The mixture is then
titrated with 0·025M potassium iodate (Section m, 123)
until the starch is completely decolourised.
I Ml. 0'025M KI0 3 == 0·001405 g. Cd.
CHAPTER IV
GRAVIMETRIC AN ALYSIS
IV, I. General discussion.-Before commencing experi-
mental work in gravimetric analysis, the student should be
familiar with the general theory underlying the chief experi-
mental processes outlined in Sections I._54 to.l, 71. He should
also read the account of the technique of gravimetric analysis
given in Sections U, 26 to U, 36; this will assume a greater
significance when the various processes have actually been
employed in practice. It is proposed, in the first place, to
give an account of a number of typical gravimetric deter-
minations. These estimations may be performed with
substances which are readily obtainable in ~ state of purity
(e.g., of analytical reagent quality), an~ the experimental
error can therefore be checked by calculation. Many may,
bowever, prefer to carry out the analyses with solutions or
solids of unknown" composition. A list of solutions and
II
of solids which are suitable for this purpose is given in the

Appendix, Section A, 7. These determinations should be
carried out before thOse described under the heading of
SYSTEMATIC GRA: VI METRIC ANALYSIS are attempted. In
general, the experimental procedures will not be given in
such detail in the latter.
IV, n. Note-book, entries, and caIculations.-A note-book
of not less than quarto size and provided with stout cardboard
covers will be found suitable for the entry of results. The
right-hand page should contain the nature of the determina-
tion, the date when it was commenced (and, if desired, com- -
pleted), a description in the past tense of the experimental
details of the estimation, together with any relevant equations.
The actual experimental results (weights, etc.), should be
entered on the left-hand page; essential details of the
calculations should also be recorded here. Finally, a line or
two should be devoted to the main results of the analysis.
Experimental results should on no account be entered on
scraps of paper, for these are readily lost.
IV, 3. Calculations of gravimetric analysis. Chemical
factors.-The calculation of the weight cif a constituent in a
given precipitate follows directly from the proportion-:
M"Ap : nM :: w : x,
'l73
where MnAp is the molecular weight of the precipitate, M
the atomic (or molecular) weight of the element (or radical)
sought, n the number of atomic (or molecular) weights of M
in the molecular weight M,.A", w is the \Yeight of precipitate,
and x is the weight of the constituent desired. Furthermore,
if W is ',the weight of the sample used, the percentage of the
constituent sought y is given by :
x : W :: y : 100,
or • • y = x X 100/W.
Example. 1·000 Gram of an iron compound, after suitable treat-
ment, yielded 0·1565 g. of ferric oxide. Calculate the percentage of
iron in the cotnpound.
Fe 20 a 2 F e : : 0·1565 : %
159·68 2 X 55·84 :: 0'1565 .. : %
(Mol. wt. of FesO.) (2 X At. Wt. of Fe)
lli.68 .
% = 159·68
-- X 0·1565
.
= 0·6994 X 0·1565
= 0·1095 g. of Fe.
Now 0·1095: 1·0000 :: y : 100,
or y = 10·95 per cent oj Fe.
Instead of evaluating proportions like the above for each
determinatioJ), we can arrive at tbe same result by one
multiplication if we mUltiply the weight of, the precipitate by
that number which represents the weight of the constituent
corresponding to 1 g. of precipitate. Such a number is
called a chemical factor. For example, the factor for Fe in
Fe 20 3 is 0·6994, and if, in the above example, we multiply
0·1565 by this number, we obtain 0·1095, which is the same
result as is obtained by evaluating the proportion.
A Table of important-chemical conversion factors, together
with their logarithms, is given in the Appendix, Section A, 2.
SIMPLE GRAVIMETRIC DETERMINATIONS

IV, 4. Determination of water of hydration in crystallised
barium chloride.-Discussion. Barium chloride dihydrate
loses all its water of crystallisation above 100°C. Much higher.
temperatures can be used in this dehydration, for anhydrous
barium chloride is non-volatile and stable even at fairly high
temperatures.
BaCl2 ,2H zO = BaC1 2 + 2H 2 0.
With some hydrated salts, special temperature limits must be
~~~~. '
Gravimetric Analysis 475
Procedure. H(:'(j.t a crucible and lid to dull redness for several
minutes, allow to cool in a desiccator, and weigh after 30
minutes. Introduce into the crncible I to 1·5 g. of A.R.
barium chloride, and weigh again. Place the covered crucible,
resting upon a pipe-clay or silica triangle, about 15 cm. above
a small flame (not more than 5 to 6 cm. high). At intervals
of a few minutes increase the flame gradually until the bottom
of the crucible is heated to dull redness. Maintain the crucible
at this temperature for about W minutes, allow it to cool in
a desiccator for 30 minutes, and weigh. Repeat the process
until constant weight (two c01:i.secutive weighings agreeing
within 0'0002 g.) is obtained. . '
From the loss in weight, calculate the percentage of water
in barium chloride dihydrate.
A number of determinations may be carried out by simple
ignition. These include iron in ferric ammonium alum
«NH4)2504,Fe2(SO,)3,24H20 _,. FezO a), aluminium in am-
monium alum (. (NH.)zS04,Al z(S04)3,24H 2 0 -+ AI 2 0 3 ),
bismuth in bismuth oxynitrate and carbonate (residue is
Bi z0 3 ), and zinc in basic zinc carbonate (residue is ZnO).
The conversion factors (see Appendix, Section A, 2) must, of
course, be known.
Note. Similar estimations may be carried out with magnesium sulphate
heptahydrate (MgS0 1• 7HP = MgSO, + 7H2 0). borax ordisodium pyroborate.
decahydrate (Na.B,07' lOH.O = Na 2 B,07 + lOHP). and with disodium
hydrogenphosphatedodecahydrate(2Na.HPO,.12H.O = Na,P.0 7 + 25H.O).
IV, 5. Determination of chloride as silver chloride.-Dis-
cltss£on. The aqueous solution of the chloride is acidified
with dilute nitric acid in order to prevent the precipitation of
other silver salts, such as the phosphate and carbonate, which
might form in neutral solution, 1.nd also to produce a more
readily filterable precipitate. A slight excess of silver nitrate
solution is added when silver chloride is precipitated:
Cl- + Ag+ -+ AgCl.
The precipitate, which is initially colloidal, is coagulated into
curds by heating the solutiQ1l and stirring the suspension
vigorously; the supernatant liquid becomes almost clear.
The precipitate is cqllected in a filter crucible, washed with
very dilute nitric acid, in order to prevent it from becoming
colloidal (Section I, 36), dried at l:lO° to 150°C, and finally
weighed a~ Agel. If a quantitative 'filter paper is used for
filtration (Section II, 28), a procedure which is not recom-
mended, the filter paper is finally washed with a little water to
476 Quaniitative Inorganic Analysis.
remove acid, which would attack the paper when dried.
The paper and precipitate is dried at 100°C, and the paper is
burnt separately (Section U t 360); the .ash is treated with a
little nitric acid and a drop or two of concentrated hydro-
chloric acid, and eyaporated on the water bath to convert the
reduced silver back to silver chloride. The main precipitate
is then added to tire crucible, and the whole dried at 130° to
150°C.
Silver chloride Jtas a solubility in water of 1·5 mg. per litre
at 21°C and 21·7 mg. per litre at 100°C. The solubility is less
in the 'Presence of very dilute nitric acid (up to 1 per cent),
and is very much less in the presence of moderate concentra-
tions of silver nitrate (see Section It 8; the optimum concen-
tration of silver nitrate is 0'05 g. per litre, but the solubility
is negligibly small up to about i'7 g. per litre). Under the
conditions of the precipitation, very little occlusion occurs.
If silver chloride is washed with pure water, it mat become
colloidal and run through the filter. For this reason the wash
solution should contain an electrolyte (c9mpare Section It 56).
Nitric acid is generally employed because it is without action
on the precipitate and is readily volatile; its concentration
need not be greater than o·OIN. Completeness of washing of
the precipitate is tested for by determining whether the excess
of the precipitating agent, silver nitrate, has been removed.
This may be done by adding 1 or 2 drops of O·lN hydrochloric
acid to 3 to 5 ml. of the washings collected after the washing
process has been continued for some time; if the solution
remains clear or exhibits only a very slight opalescence, all
the silver nitrate has been removed.
Silver chloride is light sensitive. Decomposition is only
slight in diffused light, but is considerable in bright light.
Hence the determination must be carried out in as subdued
a light as possible, and when the solution containing the
precipitate is set aside, it should be placed in the dark (e.g.,
in a locker) or the vessel containing it should be covered
with thick brown paper.
Procedure. Weigh out accurately about 0·2 g. (or an amount
containing approximately 0·1 g. of chlorine) * into a 250 tp
350 ml. beaker provided with a stirring rod and covered with
a clock glass. Add about 150 m!. of water, stir until the solid
has dissolved, a,nd add 0·5 ml. of concentrated nitric acid.
To the cold solution, add O·IN silver nitrate slowly and with
• A.R. Potassium or sodium chloride is suitable.
constant stirring. Only a slight excess should be added;
this is readily detected by allowing the precipitate to settle
and adding a_ few drops of silver nitrate solution, when no
further precipitate should be dbtained. Carry out the deter-
mination in subdued light. Heat the suspension nearly to
boiling, while stirring constantly,,· and maintain it at this
temperature until the precipitate coagulates and the super-
natant liquid is clear (2 to 3 minutes). Make certain that
precipitation is complete by adding a few drops of silver nitrate
solution to the supernatantliquid. If no further precipitate
appears, set the beaker aside in the dark, and allow the solution
to stand for at least 1 to 2 hours before filtration. In the mean-
time prepare a filter crucible (Gooch, porous porcelain or
sintered glass; the last-named is most convenient); the
crucible must be dried at the same temperature as is employed
in heating the precipitate (130 0 to 150°C) and allowed to cool
in a desiccator (see Sections fi, 31, 33 for details). Collect the
precipitate in the weighed filter crucible (Section fi, 35).
Wasli the precipitate 2 or 3 times by decantation with cold
0·01 to O·02N nitric acid before transferring the precipitate
to the crucible. Remove the last small particles of silver
chloride adhering to the beaker with a " policeman" (Section
fi, lOL). Wash the precipitate in the crucible with 0·01 to
0·02N nitric acid added in small portions (see Sections I, 60
and fi, 34) until 3 to 5' mI. of the washings, collected in a
test;-tube, give no turbidity with 1 or 2 drops of O'IN hydro-
chloric acid. * Place the crucible and contents in an air oven
at 130 0 to 150°C for I hour, allow to cool in a desiccator, and
weigh. Repeat the lreating and cooling until constant weiglit
is attained.
Calculate the percentage of chlorine in the sample.t The
factor for converting silver chloride to chlorine is 0·24737.
• A rapid method for weighinll' the silver chloride, collected in a porous
porcelain or sintered glass crucible, is as follows (J. Dick, 1929). (This method
should not be used oy elementary students or beginners in the study of quan-
titative analysis.) After washing the precipitate with very dilute nitric acid.
- wal!h the walls of the crucible 5 or 6 times with small volumes of rectified
spirit (a small pipette or a drawn-out glass tube is useful for this purpose),
followed by several times with small volumes of anhydrous ether. Suck the
precipitate dry at the pump for 10 minutes, wipe the outside of the crucible
with a clean linen cloth. leave in a vacuum desi<;cator for 10 minutes, and
weigh as Agel. The procedure may be employed for silver bromide. iodide,
and thiocyanate.
t In ·this and all other gravimetric determinations. duplicate estimations
are advisable. Both determinations may be carried out simultaneously, or
if this is not convenient, the second should be commenced as soon as pass.i.l)le
after the first is in progress.
478 Quanlt'tative Inorgant'c An.alysis
Note on the gravimetric standardisation of hydrochloric acid. The
gravimetric gtandardisation of hydrochloric aeirl hy precipitation as
silver chloride is a convenient and accurate method, which has the
additional advantage of being independent of the purity of any
primary standard (compare Section ill, 4). Measure out from a
burette 30 to 40 m!. of the, say, O'IN hydrochloric acid which is to
be standardised. Dilute to 150 mI., precipitate (but omit the addi-
tion of nitric acid), filter, and weigh the silver chlori.de. From the
weight of the precipitate, calculate the chloride concentration of the
solution, and thence the normality of the hydrochloric acid.
IV, 6. Determination of sulphate as barium sulphate.-
Discussion. , The method consists in slowly adding a dilute
solution of barium chloride to a hot solution of the sulphate
slightly acidified with hydrochloric acid:
Ba++ + S04- - -+ BaS0 4.
The precipitate is filtered off, washed with .water, carefully
ignited at a red heat, and weighed as barium sulphate. The
percentage of sutphate is calculated from the weight of barium
sulphate.
The reaction appears to be a simple one, but is in reality
subject to numerous possible errors; satisfactory results can
only be obtained if the experimental cQnditions are carefully
controlled. Before some of these are discussed, the student
is recommended to read Sections I, 56 to I, 59.
Barium sulphate has a solubility in water of about :{ mg.
per litre at the ordinary temperature. The solubility is
increased in the presence of mineral acids, because of'the
formation of the bisulphate ion (S04- - + H+ ~ HSO,-) ;
thus the solubilities at room temperature in the presence of
0·1,0·5, 1·0 and 2'ON hydrochloric acid are 10, 47, 87 and 101
mg. per litre respectively; but the solubility is less in the
presence of a moderate excess of barium ions. Nevertheless,
it is customary to carry out the precipi'tation in weakly acid
solution in order to prevent the possible formation of the
barium salts of such anions as chromate, .carbonate, and
phosphate, which are insoluble in neutral solutions; more-
ovet, the precipitate thus obtained consists of large crystq.ls
and is therefore more easily filtered (compare Section I, 57).
It is also of great importance to carry out the precipitation at
boiling temperature, for the relative supersaturation is less at
higher temperatures (compare Section I, 57). The concen-
tration of hydrochloric acid is, of course, limited by the
solubility of th~ barium sulphate, but it has been found that
a concentration .of O'05N is suitable; the solubility of the
Gravimetric Analysis 471,)
precipitate in the presence of barium chloride at this acidity
is negligible. The precipitate may be washed with cold
water, and losses, owing to solubility influences, may be
neglected except for the most accurate work.
Barium sulphate exhibits a remarkable tendency to carry
down other salts (see coprecipitation, Section I. 58). Whether
the results will be low or high will depend upon the nature of
the co precipitated salt. Thus barium chloride and barium
nitrate are readily coprecipitated. These salts will be an
addition to the true weight of the barium sulphate, hence the
results will be high, since the chloride is unchanged up.on
ignition and the nitrate .will yield barium oxide. The error
due to the chloride will be cbnsiderably reduced by the very
slow addition of hot dilute barium chloride solution to tbe hot
solution, which is constantly stirred; that due to the nitrate
cannot be ayoided, and hence nitrate ion must always be
removed by evaporation with a large excess of hydrochloric
acid before precipitation. Chlorate has.a similar effect to
nitrate, and is similarly removed.
In the presence of certain cations (sodium, potassium,
lithium. calcium. aluminium, -chromillm, and ferric iron),
coprecipitation of the sulphates of these metals occurs, and
the results will accordingly be low. This error cannot be
entirely avoided except by the· removal of the interfering
ions. Aluminium, chromium and iron may -be removed by
'precipitation, and the influence of the other ions, if present,
is reduced by considerably diluting the solution and by digest-
ing t1re precipitate (Section I, 58). It must lJe pointed out that
the general method of reprecipitation cannot be employeq. in
order to obtain a purer precipitate, because no simple solvent
(other than concentrated sulphuric acid) is available in which
the precipitate may be dissolved.
Pure barium SUlphate is not decomposed when heated in
dry air until a temperature of about 1400°C :
. BaSO, -+ BaO + S03'
The precipitate is, however, easily reduced to sulphide at
temperatures above 600°C by the car.bon of the filter paper:
B.aSO, + 4C = BaS + 4CO.
The reduction is avoided by first charring the paper without
inflaming, .and then burning off the carbon slowly !it a low
temperature with free access of air. If a reduced precipitate
is obtained, it may be re-oxidised by treatment with sulphuric
acid, followed by volatilisation of the acid and reheating.
480 Quantitative Inorganic A ttalysis
The final ignition of the barium sulphate need not be made at
a higher temperature than 600 0 to 800°C (dull red heat). A
silica Gooch crucible or a Jena quartz filter crucible may. be
used, and the difficulty of reduction by carbon is e,ntirely
avoided.
Procedure. Weigh out accurately about O·&g. of the.solid*
(or a sufficient amount to contain 0·05 to 0·1 g. of sulphur)
into a 400 ml. beaker, provided with a stirring rod and clock-
glass cover. Dissolve the solid in about 25 ml. of water, add
0·5 to 1 ml. of concentrated hydrochloric acid, and dilute to
250 to 300 ml. Heat the solution to boiling, add dropwise
from a burette or pipette 20 ml. of 5 per cent barium chloride
solution (5 g. BaCl 2,2H 20 in 10~ ml. of water-ca. 2M). Stir
the solution constantly during the addition. Allow the
precipitate to settle for a minute or two, stirring meanwhile,
and test the supernatant liquid for complete pr~cipitation by
adding a few drops of barium chloride solution. If a precipi-
tate is formed, add slowly a further 5 ml. of the reagent, and
test again; repeat this operation until an exc~ss of barium
chloride is present. When an excess of the precipitating
agent has been addeo, keep the covered solution hot, but not
boiling, for an hour (steam bath, low temperature hot plate,
or small flame) in order to allow time for complete precipi-
tation.t The volume of the solution should not be allowed to
fall below 200 ml.; if the clock glass covering the beaker is
removed, the under side must be rinsed off into the beaker
by means of a stream of water from a wash bottle. The
precipitate should settle readily and a clear supernatant
liquid should be Qbtained. Test the latter with a few drops
of parium chloride solution for complete precipitation. If
no precipitate is obtained, the barium sulphate is ready for
filtration. The determination may be completed by either
of the following processes.
(i) Filter paper method. Decant the clear solution through
an ashless filter paper (Whatman, No. 40 or No. 540) (Section
U, 28), and collect the filtrate in a clean beaker. Test the
filtrate with a. few drops of barium chloride: if a precipitate
forms, the entire sample must be discarded and a new deter-
mination commenced. If no precipitate forms', discard the
liquid, rinse out the beaker, and place it under the funnel;
this is in order to avoid the necessity of refiltering the whole
* A.R. Potassium sulphate may lJe employpd.
t An equivalent result is obtained by allowing the solulion to stand at the
ordinary temperature for about 18 hours.
Gravimetr'ic Analysis 481
solution if any precipitate should pass through the filter.
Transfer the precipitate to the filter with the aid of a jet of hot
water from the wash bottle (Section n,.35). Use a " police-
man" to remove any precipitate adhering to the walls of the
beaker or to the stirring rod, and transfer the precipitate to
the filter paper. Wash the precipitate with small portions
of hot water. Direct the jet as near the top of the filter paper
as possible, and let each portion of the wash. solution run
through before adding the next (Section n, 34). Continue
the washing until about 5 ml. of the wash solution give~ no
opalescence with a drop or two of silver nitrate solution.
Eight or ten washings are usually necessary. .
Fold the mois~ paper around the precipitate and place it in a
weighed porcelain crucible, previously ignited to redness and
cooled in a desiccator. Dry the paper by placing the loosely:
covered crucible upon a triangle several ems. above a small
flame (Section n, 36). Then grq.dually increase the heat until
the paper chars and volatile matter is expelled. Do not
allow the paper to burst into flame as mechanical loss may
thus ensue. When the charring is complete, raise the terp:
perature of the crucible to dull redness, and burn off the carbon
with free access of air* (crucible slightly inclined with cover
displaced, Fig. 2-37). When the precipitate is white, ignite
the crucible at a red heat for 10 to 15 minutes. Then allow
the crucible to cool somewhat in the air, transfer it to a
desiccator, and, when cold, weigh the crucible and contents.
Repeat the ignition with 10 minute periods of heating, sub-
sequent cooling in a desiccator, etc., until constant weight
(± 0·0002 g.) is attained.
Calculate the percentage of SO, in the sample. The
factor for converting BaSO, to S is 0·13735.
(ii). Gooch crucible method. Prepare a Gooch crucible as
described in Section n, 31. A silica Gooch crucible is pre-
ferable to one 'of porcelain; alternatively a Jena quartz filter
crucible (Section n, 33) may be used, and no preliminary pre-
paration, apart from drying at the correct temperature, is
necessary. Ignite this inside a porcelain or nickel crucible
(Fig. 2-33) at a red heat (or in an electric muffle furnace at
600 0 to 800 oq, allow to coo~ in a desiccator, and weigh.
Filter the supernatant liquid. after digestion of the precipi-
tate, through the weighed crucible, using gentle suction.
Reject the filtrate, after testing for complete precipitation
* Any dark matter on thll crucible cover may be removed by placing it, •
clean side down, ou a triangle, and heating it for some time.
R
482 Quantitative inorganic Analysis
with a little barium chloride solution. Transfer the precipi-
tate to the crucible ana wash with warm water until 5 ml. of
the filtrate gives no turbidity with a few drops of silver nitrate
solution. Dry the crucible and barium sulphate"'in an oven,
and thep ignite in a manner similar to that used for the empty
crucible for periods of 15 minutes until constant weight is
attained (1).
Note. 1. A rapid method for weighing the precipitate is as follows (J.
Dick, 1929). (This procedure should not be employed by elementary students
or beginners in the study of quantitative analysis.) ;Filter off the precipitated
barium sulphate through a weighed crucible (Gooch, sintered glass or porous
porcelain) and wash it with hot water until the chloride reaction of the wash-
ings is negative. Then wash 5. or 6 times with small volumes of rectified
spirit, followed by 5 timd with small volumes of anhydrous ether. Suck the
precipitate dry on the pump for 10 minutes, wipe the outside of the crucible
dry with a clean linen cloth, leave in a vacuum desiccator for 10 minutcs (or
until constant in weight), and weigh as BaSO,.
Calculate the percentage of S04 in the sample.

IV. 7. Determination of sulphur in iron pyrites.- Disc1tssion.
The methods to be described apply to' most insoluble sul-
p,hides. In these the sulphur is oxidised tcysulphuric acid,
and determined as barium sulphate. Two procedures are
available for effecting the oxidation.
A. Dry Process. The oxidation is carried out by fusion
with sodium peroxide, or, less efficiently, with sodium car-
bonate.and potassi'um nitrate.
2FeS2 + 15Na 0 =2 2 Fe 20 3 + 4Na 2S04 + llNa 20.
The sulphide is fused in an iron or nickel crucible with the
sodium peroxide (platinum is strongly attacked-Sections
n. 16 and n. 37), the fused mass treated with water, filtered
and acidified. The excess of peroxide is removed by boiling,
and the sulphate ion precipitated with barium c~loride .. The
decomposition of the sulphide is rapid, but the method has
several disadvantages. Amongst these may be mentioned:
the slight attack on the metal crucible, thus preventing the
subsequent determination of the me~l content of the sample;
the introduction of appreciable quantities of sodium salts,
thus increasing the error due to coprecipitation (Sections
I. 58 an<1 IV, 6); and the danger of contamination of the
sulphur from the flame gases, since sulphur dioxide is rapidly
absorbed by the alkaline melt. The last error may be mini-
mis~d by fitting the crucible into a hole in a sheet of asbestos
• or .. uralite," and keeping the crucible covered during the
ignition (see Section II, 38).
B. Wet Process. The sulphide is oxidised (i) by bromine in
carbon tetrachloride solution, followed by nitric acid, (ii) by
sodium chlorate and hydrochloric acid, or (iii) by a mixture
of nitric and hydrochloric acids and a little _bromine. The
use of the last-named oxidising agent will be described; the
reaction may be represented by : ,
2FeS z + 6HNO s + 15Br2 + 15H 2 0
= 2Fe(NO a)s + 4H zS0 4 + 30HBr.
The method has the advantage of not introducing any metallic
ions, but it is essel1tial to remove the excess of nitric acid
(see Discussion in Section IV, 6). The action is slower than
by the fusion method.
Procedure. A. Dry Process. Dry some finely-powdered
pyrites* at 100°C for 1 hour. Fit an iron or nickel crucible
into a hole in' an asbestos or uralite" board sufficiently
C(
large to allow two-thirds of the crucible to project below the

board. Place about 1 g . .of A.R. anhydrous sodium carbonate
into the crucible, and weigh accuratel)L into it 0·4 to 0·5 g. of
the pyrites. Add 5 to 6 g. of sodium peroxide, and mix well
with a stout copper or nickel wire or with a thin glass rod.
Wipe ~he wire or rod, if necessary, with a small piece of
quanfitative filt~r' paper, and add the latter to the crucible;
cover the mixture with a thin layer of peroxide. Place the
crucible in the hole in the asbestos- or " uralite " sheet, and
heat it with 'a very small flame, Increase the temperature
gradually until after 10 to 15 minutes the crucible is at a dull
red heat (the lower the temperature, the 'less is the cfUcible
attacked) and just sufficient to keep the .mass completely
fused. Remove the cover occasionally and examine the
contents; be sure that 'the whole mass is fluid. Maintain
the mass fluid for 15 minutes to complete the oxidation.
Allow to cool, extract the crucible with water in a covered
600 m!. beaker, rinse off the crucible cover into the beaker,
remove the crucible with a glass rod and wash it well; dilute
to 300 m!. Boil the solution for 15 minutes in order to,destroy
the excess of peroxide, add a What man '~accelerator" or a
quarter of an ashless tablet," and filter through a Whatman
C(
No. 54:1 filter ,paper. Wash the residue at least 10 times

with hot 1 per cent sodium carbonate solution. Acidify
the combined filtrate anq washings contained in a 800 to
1000 m!. beaker with concentrated hydroclJloric acid, u~ng
.. Ridsdale's " Iron Pyrites, No. 28" (one of the Analysed Samples for
Students) is suitable.
methyl red or methyl orange as indicator, atld add 2 ml. of
acid in excess. Dilute, if necessary, to 600 mI., and heat to
boiling. Precipitate the sulphuric acid by the slow addition
with stirring of a boiling 5 per cent solution of barium chloride;
the latter is added in slight excess of the calculated amount
required, assuming the pyrites to be pure FeS z• Complete
the determination as in Section IV. 6.
Calculate the percentage of sulphur in the sample.
B. Wet process. Weigh ou.t accurately about 0·5 g. of the
finely-powd"ered pyrites into a 250 m!. beaker, and pour on to
it a mixture of 3 volumes of concentrated nitric acid and 1
volume "of concentrated hydrochloric acid to which 1 ml. of
pure, sulphur-free-liquid bromine (e.g., A.R.) has been added.
Cover the beaker with a well-fitting clock glass, let the reaction
proceed at room temperature for 30 minutes, and then transfer
the beaker to a steam bath. Heat gently until.all the appar-
ent action has subsided, raise the cover,~ and evaporate the
liquid to dryness. Treat the residue with 5 -ro1. of concen--
trated hydrochloric acid and again evaporate to dryn~ss.
Moisten the dry residue with 2 ml. of concentrated hydro-
chloric acid, add 100 m!. of hot water, filter into a 400 m1.
beaker, and wash the residue with hot water. "If the residue
is small, the filtration may be omitted. TO" the hot filtrate,
which should have a volume of 200 to 250 m1., add 1: 1
ammonia solution slowly and with constant stirring until the
solution is slightly ammoniacal, and add a further 3 to 5 m!.
Filter off the ferric hydroxide (Section IV. 8) and wash it with
hot water until free from chloride; preserve the filtrate (a).
Dissolve the precipitate in dilute hydrochloric atid, repre-
cipitate the ferric hydroxide as before, filter, and add the
filtrate and washings to the one previously obtained (a).
(This double precipitation is necessary to remove the copre-
cipitated sulphate ions from the ferric hydroxide.) Render
the combined filtrates and washings just acid with hydro-
chloric acid, using methyl orange as indicator, add 2 J:!ll. of
concentrated hydrochloric acid in excess, dilute to 500 mI.,
and .precipitate the sulphuric acid as described in the Dry
Process above.
Calculate the percentage of sulphur in the sample.
IV, 8. Determination of iron as ferric oxide.-Discussioll.
The solution containing the ferric saltt is treated with a
• This is unnecessary if a. Fisher" speedyvap .. beaker cover is used.
t Ferrous iron is only partially precipitated by ammonia solution in the
presence of ammonium salts.
Gravi1lletric A Ila/vsis 485
slight excess of ammonium hydroxide solution to precipitate
ferric h ydroxide* :
FeCl 3 + 3NH.OH = Fe(OH)a + 3NH,Cl.
Other clements. that are precipitated by ammonia solution
must, of course, be absent. These include aluminium, tri-
valent chromium, titanium and zirconium. In the presence
of an oxidising agent (even atmospheric oxygen), manganese
may be precipitated as the hydrated dioxide. Anions, such
as arsenate, phosphate, vanadate and silicate, which yield
insoluble .compounds of iron in weakly basic media, must be
absent. The presence of a sufficient quantity uf salts of
organic hydroxy-acids (e.g., citric, tartaric and salicylic
acids), hydroxy compounds (e.g., 'glycerine and sugars),
alkali pyrophosphate and fluorides must be guarded against,
because of the formation of complex salts and the consequent
non-precipitation of ferric hydroxide.
The solubility product of ferric hydroxiue is of the order of
10-36 , so that quantitative precipitation occurs even in weakly
acid solution, and errors due to washing will be negligibly
small. The precipitate first forms as a dispersed phase, but
on heating in the presence of electrolytes it coagulates to a
gelatinous mass, wpich settles out of suspension; prolonged
heating tends-to break -up the aggregates and causes the pre-
cipitate to become slimy. The hydrated ferric oxide is a
typical example of a flocculated colloid. The coagulation of a
colloidal precipitate, and especially the agglomeration of the
nrimary particles, is aided considerably by raising the tem-
perature of the solution. Hence precipitation is carrieu out
at or ncar the boiling point, and the liquid is maintained
at this temperature for a short time after precipitation.
As might be expected from its colloidal character, hydrated
ferric oxide has a great tendency to adsorb other ions present.
If precipitation' is made 1rom basic solution, the primary
adsorbed ion is the hydroxyl ion (Section I, 56), and this
readily holds by secondary adsorption positive ions wl~ich
may be present. If there is a large excess of ammonium ion:>
in the precipitating and wash solutions, the agsorption Of
other cations can be kept at a minjmum; since ammonium
* The precipitate is usually given the formula Fe(OH} •• ferric hydroxide.
Actually the water content is somewhat indefinite; the formula of the
precipitate is more accurately represented by Fe.0 3 ••tl[.O. and is termed a
hydrated oxide. Throughout this book the terms hydrated oxide and hydroxide
will be t_lsed syn'Ouymously.
2Fc H + + GNlJ,OH + yHaO.:= Fe.O •• xH.O + 6N1I4+.
486 Qnanit'tafive I noyganic Analysis
salts are volatilised upon ignition of the precipitate, little
h",rm is caused by the adsorptio~. Divalent ions are more
strongly adsorbed than monovalent ions (Section I, 58). If
the extent of co precipitation is large, reprecipitation may be
employed siItce the precipitate is soluble in dilute acids.,
The gelatinous precipitate of ferric hydroxide is always
filtered through filter paper. Filtration with suction, in
order to hasten filttation, should not be used; since the effect
of the suction is merely to force the small particles of the
precipitate into the pores of the filtering medium. It may
often happen that with suction the liquid will pass through
more rapidly; this does not mean that the washing process is
accelerated since the li<Luid runs through small channels and
does not permeate the main body of the precipitate. For
this reason ferric hydroxide is best wa~ed by decantation;
the pr~cipitate may then be thoroughly stirr~d with the wash
liquid. To prevent peptis at ion and the production of slimy
material, an electrolyte is used in the wash li~uid. The
most satisfactory is ammonium nitrate; ~his volattlises upon
ignition and assists somewhat in the subsequent ignition of the
precipitate. Ammonium chloride is unsuitable, because
ferric chloride, which is volatile, is formed during the ignition-:
FelPs + 6NH4Cl = 2FeCI a +, 6NH3 + 3H zO.
It is advisable, theref<1re, to wash out nearly all the ammo-
nium chloride present in the hydrated ferric oxide: very small
amounts, however, will not lead to any significant error. To
assist filtration, a hot wash solution should be employed ..
Tlie filtration and washing af any gelatinous precipitate is
.hastened by the use of ashless filter p'aper pulp (macerated
filter paper, Section n, 29). Under no circumstances should
the. precipitate be allowed to stand in the filter paper before
washing is cO:qlplete, because it shrinks rapidly as it pl.lrtially
dries, and channels, which permit the wash liquid to run
through, are formed in the precipitate.
Hydrated ferric oxide upon ignition at lOOO°C yields ferric
o;:cide; at higher temperatures triferric tetroxide is slowly
formed.- The ignition should be carried out under good
oxidising conditions especially during the burning of the filter
paper, for otherwise partial reduction to the magnetic oxide
FeaO(, or even to the metal, may occur. These reduction
products are only slowly converted into ferric oxide upon
continued heating with free access of air. Such reduction is
avoided by burning off the carbon at a low 'heat, by main-
taining at all times free access of air, and by, excluding the
reducing gases from the flame.
Procedure. For practice in this determination, the student
may employ either A.R. ferrous ammonium sulphate
FeSO,,(NH4)2S0,,6H20 or A.R. ferric ammonium sulphate
(NH4) 2S0 4,Fe 2(SO 4) 3,24H 20. * The former is to be preferred
as this estimation involves oxidation of the ferrous salt to the
ferric state. Weigh out accurately about 0-8 g. of ferrous
ammonium sulphate into a 400 ml. beaketlprovided with a
Clock glass and stirring rod. Dis~olve it in 50 ml. of water.
and 10 ml. of dilute hydrochloric :rcid (1 : 1). Add 1 to 2 ml.
of concentrated nitric acidt to the solution, and boil gently
until the colour is clear yellow (3 to 5 minutes is usually
necessary) (1). Dilute the solutioll to 200 mq, heat to
boiling, and slowly add pure 1: 1 ammonia solution (2) in a
slow stream from a small beaker until a slight excess is present,
as.is shown by the odour of the steam above the liquid (3).
Boil the liquid gently for 1 minute, and allow the precipitate
to settle. The supernatant liquid should be colourless. As
soon as most of the precipitate has settled, decant the super-
natant liquid through an ashless filter paper, but leave as much
of the precipitate as possible in the beaker. It is essential
that the filter paper fit the fuimel properly (Section II, 28),
so that the stem of the funnel is always filled with liquid,
otherwise filtration will be very slow. A,dd about 100 ml. of
boiling 1 per cent ammonium nitrate solution to the precipi-
tate, stir the mixture thoroughly, and allow to settle. Decant
as much liquid as possible through the fi:lter. Wash the
precipitate three ta four times by decantation with 75 to 100
ml. portions of hot 1 per cent ammonium nitrate solution.
Transfer the precipitate (and macerated filter paper, if
employed) to the filter (Section II; 35); any small particles
adhering to the sides of the vessel or to the glass rod are dis-
lodged with the aid of a "policeman," and subsequently
transferred to the main precipitate with the assistance of hot
water from a wash bottle: Wash the ferric hydroxide several
times with hot ammonium nitrate solution (4) until no test
* Alternatiyely, the sample supplied by the teacher may be used.
t Alternatively. bromine water or hydrogen peroxide may be used for the
oxidation. The latter is conveniently employed as the" 100 volume" solu-
tion; 1 ml. of this solution is sufficient. The excess of the reagent must be
destroyed by boiling the solution.
t If iron alum is used. dissolve 1·3 g. (or a sufficient amount of a ferric salt
containing about 0·15 g. of iron) in 200 ml. of water, add 10 ml. of 1 : 1 hydro-
chloric acid, and proceed as described.
4SR Quantitaillle Illorganic A l1alysis
(or, at most, a very slight test) for chlorid.e is outained from
the washings. Allow each portion of the wash liquid to run
through before adding the next portion; do not fill the filter
more than thr~e-fol1rtbs full of the precipitate. While, the
filtration is in progress, ignite a clean crucible (porcelain, silica,
or platinum) at a red heat, cool in a desiccator for ao minutes,
alld weigh. When 'the filter paper has drained thoroughly,
fold over the. edges, and transfer to the weighed crucible (1)).
Proceed as described in Section IV. 6. Heat gradually until
dry, char the paper without inflaming, and burn off thc carbon
as at a Iowa temperature- as possible under good oxidising
conditions. Fina!ly, ignite the precipita!e at a red heat for
15 minutes 'and take care to exclude the flame gases from the
interior of the crucible, cool in a desiccator for 30 minutes, and
weigh. Repeat the ignition (10 to 15 minutes) until constant
weight is obtained (to within 0·0002 g.).
From the weight of ferric oxide obtained, calculate the
percentage of iron in the salt used. The factor for conv<:!rt-
ing Fe 20 3 to Fe is 0.69940.
Notes. 1. At this stage it is advisable to test the solution for the complctc
oxidation of the iron. Transfer a drop of the solution to a test-tubc by means
of a stirring rod, and dilutc with about I ml. of water. Add a few drops of a
freshly-prepared potassium ferricyanide solution. If a blue colour appcars,
ferrous iron is still present in the solution, and morc nitric acid must be
added.
2. Filtered ammonia solution should be used in ordar to prevent tJ1C intro-
duction of silica, which is often present in suspf'nsion in alkaline solutions.
:r. At this point it is actvantagf'olls to add a little macerat!'d lilter paper.
Lcst in the form of a Whatman .. accelcrator" or "ashless tablet;." Fur
further details, see Section II, 29.
4. I f desired, hot wat!'r from a wash bottle may be substituted at this stage;
l'eptisation is negligiblc.
5. The most convenicnt cruciIJ]e is a silica crucible with an interim-fitting
serrate,] lid (I\'lain-Smith type). S('e Section II, 'iIB.
For other methods for the determination of iron, see Section

IV. 31.
,.
IV. 9. Determination of aluminium. as aluminium oxide.-
Discussion. The aluminium is precipitated as the hydrated
oxide by means of ammonia solution in the presence of
ammonium chloride. The gelatinous precipitate is washed,
converted into the oxide by ignition, and weighed as A1 2 0 3 •
This de~ermination is subject to several sources of error,
most of which will now be discu?sed.. Aluminium hydroxide
is amphoteric in character:
Al(OH)3 3Ht + Al t-H + :lH 20;
AI(OH}:r + 011- , A10 2- + 2H 20.
GrCl1Jimt'iril' A na[.":;/ s 489
Precipitation commences at approximately pH 4, and is com-
plete when the pH lies between G·5 and 7·;;. The latter pH
range can be maintained by the use of methyl red as indicator.
The amonnt of ammonia solution employed for precipitation
must clearly be ~ontrollcd. This is achie-vcd by thc addition
of ammonium chloride, which exerts a buffering eHe'ct
(Section I; 20) and also assists the coagulation of the initially
colloidal rrecipitate. The presence of ammonium salts
reduces to a minimum the coprecipitation of thc divalent
metals, such as calcium and magnesium (see Scction IV, 8)
and other cations. A readily filterable precipitate is obtainec!
by precipitation in hot solution. The precipitate cannot be
washed with hot water for aluminium hydroxide is readily
peptised (Secti9n I, 56), and will run through the filter. .A
two per cent solution of either ammonium chloride or ammo-
nium nitratc is satisfactory; thc presence of amnlonillm
chloride in thc precipitate causes no appreciable volatilisa-
tion of aluminium during the subsequent ignition (contrast
ferric oxide).
The aluminium oxide obtained by igniting 'aluminium
hydroxide is hygroscopic unless the temperaturc has been
raised to. at least 1200°C, when apparently a non-hygroscopic
form of the oxide isjormed. For this reason the precipitate
is best ignited in a. silica crucible (porcelain is slightly hygro-
scopic when heated to a high temperature) over a Mckcr or
Fisher-burner or with a blast lamp.
Procedure. Weigh out accurately about 1'8 g. of A.R
ammonium aluminium sulphate (NH4)2S0"Al2(S04)3,24H20
(or a weight of a sample containing about 0·1 g. of aluminium)
into a 400 or 600 mI. beaker, provided with a clock-glass
cover and a stirring rod. Dissolve it in 200 m!. of water,
add 5 g. of pure ammonium chloride, a few drops of methyl
red indicator (0·2 per cent alcoholic solution), and heat just
to boiling. Add pure dilute ammonia solutioll. (1: 1) drop-
wise from a burette until the colour of the solution changcs to
a distinct yellow. Boil the solution for 1 .or 2 minutes, and
filter at once through a suitable quantitative filter paper
(Section II, 28) (1). Wash the precipitate thoroughly with
hot 2 per cent ammonium nitrate or chloride solution. Place
the paper with the precipitate in a previously ignited silica or
platinum crucible (2), dry, char, and ignite for IOta 15 minutes
with a Meker or Fisher high temperature burner. Allow the
crucible to cool in a desiccator, and weigh. Ignite to'constant
weight.
Dill
490 Q1eantt'tative lnorga'nic Analysis
Calculate the percentage of aluminium in the sample. The
factor for converting Al z0 3 to Al is 0·52913.
Notes.!. Macerated filter paper (Section n, 29) may be added in order to
assist the subsequent filtration,
2. A silica crucible of the Main-Smith type (Section n, 36B) considerably
simplifies this operation .•
For other methods for the determination of aluminium, see
Section IV, 28.
IV, 10. Determination of calcium as oxalate.-Discussion.
The calcium is precipitated as calcium oxalate CaC 2 0.,H 2 0
by treating a hot hydrochloric acid solution with ammonium
<1xalate, and slowly neutralising with ammonium hydroxide
solution:
Ca++ + C2 0 4 - - + H 2 0 -+ CaC 2 0 4 ,H 2 0.
Tfie precipitate is washed with dilute ammonium oxalate
solution and then weighed in one .of the following forins :
(i) as CaC20 o H zO by drying at 100 0 to i15°C for 1 to 2
hours. This method is not recommended for accjlrate work,
because inter alia of the hygroscopic nature of the oxalate and
the difficulty of removing the coprecipitated ammonium
oxalate at this low temperature.
(ii) As CaC0 3 by drying at 475 to 525°C in an electric
0
muffle furnace. This is the most satisfactory metliod sil).ce

calcium carbonate is non-hygroscopic. . . .
CaC 2 0, = CaCO a + CO 2 ,
(iii) As CaO by igniting at 1200°C. This.method is-widely
used, but the resultant lime has a.. cdmparatively small
molecular weight and is hygroscopic; precautions must
therefore be taken to prevent absorption of moisture (and of
carbon dioxide).
CaC0 3 + CaO + CO 2 ,
Calcium oxalate monohydrate has a solubility of 0·0067 g.
and 0·0140 g. per litre at 25 and 95°C respectively. The
0
solubility is -less in neutral solutions containing moderate

concen1;rations of ammonium oxalate owing to the common
ion effect (Section I, 9) ; hence a dilute solution of ammonium
oxalate is employed as the wash liquid in the gravimetric
determination. Calcium oxalate being the salt of a w~ak acid;
its solubility increases with increasing hydrogen ion concen-
tration of the solution because of the removal of the oxalate
ions (compare Section I, 13) to form the weak bi-oxalate ions
and oxalic acid:
CaC 2 0. (solid) ~ Ca++ + C20.- - ;
C2 0 4- - + H+ ~ HC 20,-; HC 2 0,- + H+ ~ HsCzO,.
Calculation shows that precipitation 'is quantitative at a pH
of 4, or higher.
Precip_itation from cold neutral or ammoniacal solutions
yields a very finely-divided precipitate, which is difficult to
filter. The most satisfactory results are obtained by adding
ammonium oxalate to a hot acid solution of the calcium salt
(more Qf less calcium oxalate may precipitate, depending
upon the pH of the solution), and finally neutralising with
ammonium hydroxide solution. The precipitate thus ob-
tained consists of relatively coarse crystals which are readily
filtered.
tn this determination a~l those metals (e.g., copper, lead,
zinc) which form slightly soluble oxalates must be absent.
The problem which frequently arises in practice is the pre-
cipitation of calcium in the presence of magnesium and the
alkali metals. The amount of the alkali metals which is
precipitated is usually small; III the presence of large amounts
of sodium,'reprecipitation may be desirable. Magnesium may
be coprecipitated (Section It 58) to a considerable exte:'1t, but
the amount of this may be considerably reduced by not boilin,g
the solution and not allowing the precipitate to stand in
contact with the solution too long before filtration (post-
precipitation, Section It 58, is thus minimised). By using a
very large excess of ammonium oxalate, magnesium is held
in solution in the form of a complex salt with .the oxalate
ion; furthermore, magnesium oxalate readily forms quite
stable supersaturated solutions. -If the concentration ratio
of magnesium to calcium is extremely large, a second pre-
cipitation may be necessary.
As already pointed out, the two forms in which calcium
oxalate is best weighed are the carbonate and ox\de. The
theory of the decomposition of calcium oxalate is of '>ome
interest in this connexion. Decomposition of the oxalate
,into the carbonate is rapid at about 475°C. At higher.
temperatures, the dissociation of calcium carbonate
(CaCO s ~ CaO + C~) comes into play. At any given temper-
ature, a mixture of CaC0 3 , Caq and CO 2 in equilibrium with
one another exerts a certain definite pressure of carbon diox-
ide. If the partial pressure of the carbon dioxide in the sur-
rounding ,atmosphere is greater than the equilibrium pressure
for tliat temperature, the above reaction will proceed from
right to left, and eventually the oxide wilf be completely con-
verted into the carbonate. Otherwise expressed, calcium car-
bonate cannot be decomposed into the oxide so long as the
pressure'of carbon dioxide in the surrounding atmosphere is
greater than the equilibrium pressure of the system
CaCO a-CaO-C0 2 at the temperature of heating. Atmospheric
air contains about 0·03 per cent of carbon dioxide by
volume; when the pressure is 760 mm., this corresponds to
,760 X 0·0003 = 0·228 mm. of mercury. Calcium carbonate
will therefore be perfectly stable in the atmosphere- so long
as the decomposition pressure does not exceed 0·23 mm. of
mercury. The dissoGiation pressure of calcium carbonate,
expressed, in mm, of mercury, at various temperatures are
collected
.
. in the following table. .
T~mp. Dissociation pressureJ T('mp . Di.$sociation pressure.
._-_'.-
:!IlOnc 7·8 X 10- 9 700°C 3l·2
400°C 0·3 X 10--3 800° 208
:'00° 0·15 882° 76U
liOO° 2·98 900 0 984
I
Thus calcium carbonate will not commence to dissociate

appreciably in the atmosphere until a temperature of slightly
above I)OO°C. Actual experiment has shown that complete
decomposition of calcium oxalate into the carbonate occurs
at a temper;ature between 475° and 525°C; above 550°C the
calcium carbonate commences to lose carbon dioxide. For
the weighing of calcium oxalate as calcium carbonate, fine
temperature control is necessary; this can only be achieved
by the use of an electrically-heated muffle furnace, provided
.with a pyrometer or suitable thermometer. If such equip-
ment is available, the method should be used in preference to
all others; the oxalate must be filtered through a Gooch
(preferably of silica) or porcelain filtering crucible and not
• through filter paper.
Above 88-2°C calcium carbonate is completely decomposed
into the oxide, but unless the carbon dioxide is removed by
diffusion: convection, etc'. , b:y conducting the ignition in a
loosely-covered crucible, there will be a recombination of
calcium oxide and carbon dioxide on cooling with the forma-
tion of some calcium carbonate. In practice it is ~ound that
the rate of decomp.osition at about 900°C is very slow; and it
is best to use a temperature of 1l00° to 1200°C. This tem-
perature is not easily attained in a porc~lain or silica crucible
unless an electrically heated muffle furnace is employed.
'Gravimclric Analys£s
However, quantities up to 1 g. can be completely decomposed
in a platinum crucible by the use of a Meker or Fisher high
temperature burner. The residue of calcium oxide is hygro-
scopic (unless heated for a considerable time above 1200°C).
The crucible should be kept covered in a desicc;,ttor, containing
pure concentrated sulphuric acid, freshly ignited quicklime,
or stick sodium hydroxide, for only 25 to 30 minutes, and
weighed,.
Procedure. Weigh out accurately sufficient of the sample to
.contain 0·2 g. of calcium* into a covered 400 or 600 m!.
beaker provided with a stirring rod. Add 10 m!. of waUr,
followed by about 15 m!. of dilute hydrochloric acid (1 : 1):
Heat the mixture until the solid has dissolved, and boil
gently for several minutes in order to expel carbon dioxide.
Rinse down the sides of the beaker and the clock glass, and
dilute \0200 m!.: add 2 drops of metliyl red indicator. Heat
the solution to boiling, and add very slowly a warm solution
of 2 g. of ammonium oxalate in 50 ml. of water. Add to the
resultant solution filtered dilute ammonia solution (1: 1)
dropwise and with stirring until the mixture is neutral or
faintly alkaline (colour change from red to yellow). Allow
the solutidll to stand without further heating for at least an
hour. After the precipitate has settled, test the solution
with a few' drops of ammonium oxalate solution for'complete
precipitation. The subsequent procedure will depend as to
whether the calcium oxalate is to be weighed as the carbonate
or as the oxide.
Weighing as calcium carbonate. Decant the clear supernat-
ant liquid through a weighed silica Gooch crucible or a porous
porcelain filtering crucible. Transfer the. precipitate to the
crucible with a jet of water from the wash bottle; any preci-
pitate adhering to the beaker or to the stirring rod is trans-
-ferred witli the aid of a rubber-tipped rod (" policeman ").
Wash the precipitate with a cold, very dilute ammonium
oxalate'solution (0:1 to 0,2 per cent) at least 5 times, or until
the washings give no test for chloride ion (add dilute nitric
acid and a few drops of silver nitrate solution to 5 m!. of the
washings). Dry the precipitate in the steam oven or at HO°
to 120°C for 1 hour, and then fransfer to an electrically-
heated muffle furnace, maintained at '500°C, for 2 hours.
Cool the crucibie and contents in a desiccator, and weigh.
• 0'5 Gram of A.R. calcium carbonate, or of calcite, which has been finely
powdered in an agate mortar and dried at 110° to 130°C for 1 hour, is suitable.
Further heating at 500°C should not affect the weight. As a
final precaution, moisten th~ precipitate with a few drops of
saturated ammonium carbonate solution, dry at 110°C, and
weigh again. A gain in weight indicates that some oxide was
present; this should not occur.
Calculate the percentage of calcium in the sample. The
factor for converting CaCO a to Ca is 0·40044.
Weighing as calcium oxide. Decant the clear supernatant
liquid through a Whatman No. 40 or 540 filter paper, transfer
the precipitate to the filter (Section fi, 35), and wash with t
cold 0·1 to 0·2 per cent ammonium oxalate solution until free
from chloride. Transfer the moist precipitate to a previously
ignited and weighed platinum crucible, and ignite gently at
first over a Bunsen flame and finally for !O to 15 minutes with
a Mcker or Fisher hig'R temperature burner until two succes-
sive w~ighings do not differ b'y more than 0,0003 g. The
crucible and contents are weighed after cooling in a desiccator
containing pure concentrated sulphuric acid or stick sodium
hydroxide (but not calcium chloride) for 25 minutes only.
Calculate the percentage of calcium in the sample. The
factor for converting CaO to Ca is 0·71469. .,
For other methods for the determination of calcium, see
Section:W,43.
IV, 11. Determination of magnesium as pyrophosph~te.

Discussion. An acid solution of the magnesium salt is treated
with an excess of diammonium hydrogen phosphate, and then
excess of ammonia solution is added to precipitate magnesium
ammonium phosphate at room temperature:
Mg++ + NH,+ + ~04- - - + 6H 2 0 = MgNH 4P0 4 ,6H 20.
The precipitate is wasb£d with dilute ammonia solution, *
ignited to magneslum pyrophosphate at a high temperature
(1l00°C), and weighed as such.
2MgNH 4P0 4 = Mg 2P 20 7 + 2NH.3 + 2H 20 ••
To obtain a precipitate of correct cO'mposition (MgNH4P04)
at the first precipitation is a difficult matter owing to the
·Results of moderate accuracy may be obtained by weighing the precipitate
as the MgNH4PO •. 6Hp. After washing on the filter qucible with dilute
ammonia solution to remove soluble salts. the precipitate is washed with four
2-4 ml. portions of rectified spirit. followed by several times with small volumes
of n.nhydrou~ ether. The precipitate is sucked dry at the ·pump for 10
minuteS', the outside of the crucible wiped with a clean linen cloth, a.,nd left in
a vacuum desiccator for 10-20 minutes; the llrecipitate is then weighed as
l\1gNH. PO.,6lIp.
coprecipitation of ammonium phosphate and magnesium
phosphates; however, if the experimental conditions are
carefully chosen and a pure magnesii'lm salt is used, a pre-
cipitate of normal composition is formed. If the precipitation
takes place in the presence of much ammonium salts, the
precipitate contaiI1s Mg 3 (P0 4)2 or Mg(NH 4)4(P0 4)2; the
Lormer is unchanged by ignition, and the latter gives magne-'
sium metaphosphate. If the precipitation is made 'n the
presence of muc~ potassium or sodium salts, the precipitate
is contaminated with magnesium potassium (or sodium)
phosphate. Hence if much ammonium, potassium or so(iium
salts are p,resent, reprecipitation is essential. In any case,
reprecipitation is desirable to procure the best results. The
double precipitation process will accordingly be described.
It is an experimental fact that the precipitate is practically
insoluble in 5 per cent ammonia solution: this is accordingly
used as the wash liquid.
Great care must be taken in the conversion of magnesium
ammonium phosphate into the pyrophosphate. The carbon
must be burnt off at as lo~ a temperature as possible,
because of the .danger· of the reduction of the phosphate
precipitate if the heating is strong while carbon remains; if a
platinum crucible is used, the resultant phosphorus may lead
to serious injury of the crucible. Furthermore, if the heating
is rapid, a dark coloured product is obtained. For these
reasons, the charring of the paper and the burning off of the
'carbon are conducted at as low a temperature as possible;
the temperature must be raised very gradually. Some
authors recommend, particularly for elementary students,
that the filter paper be ignited apart from the precipitate
(Section n, 36C) ill order'to minimise this danger. A porce-
lain filter crucible may be used to collect the precipitate; this
is then heated in an electric muffle furnace at 1050"
to 1l00° C.
Procedure. To a neutral or slightly acid (hydrochloric)
solution of a magnesium compound,·containing not more than
0·1 g. of magnesium, * add 5 ml. of concentrated hydrochloric
acid, and dilute. to 150 ml. Add a few drops of methyl red
indicator to the cold solution, and then 10 ml. of the freshly-
prepared ammonium phosphate reagent (2ii g. of A.R
(NH4)2HP04 dissolved in 100 ml. of water). Now add pure
• About 0·6 gram of A.R. magnesium sulphate. accurately weighed. i~ a
convenient quantity of magnc~iulll salt to elllploy for this estimation.
concentratp.d ammonia solution slowly while stirring the solu-
tion vigorously until the indicator turns yellow. Avoid
scratching the sides of 'the beaker with the stirring rod, for
wherever there is contact, an adhering crystalline deposit
forms quickly. Cdntinue to stir the solution for 5 minutes,
adding ammonia solution dropwise to keep the solution
'Yellow, and firially add {) ml. of concentrated ammonia
solution in excess. Allow the solution to stand. in a cool
place for at least 4 hours or preferably overnight. Filter
through a quantitative filter paper withoul attempting to
transfer the precipitate completely, and wash the ve~sel,
precipitate and paper several times with cold 1 : 20 ammonium
hydroxide solution. Dissolve the precipitate on the paper in
approximately 50 ml. of warm dilute hydrochloric acid (1: 10),
and wash the paper through with hot very dilute hydro-
chloric acid (1 : 100) into the b~aker used for the initial
precipitation, dilute to 100 to 150 mI., add a few drops of
methyl red indicator, 1 ml. of the ammonium phosphate
reagent, and cool to room temperature or, better, in ice.
Complete the precipitation by the addition of concentratt;d
ammonia solution dropwise and with' constant stirring until
the solution is yellow, and then add 5 ml. in excess. Let the
solution stand for at least 4 hours or, better, overnight. Filjer
the precipitate through a quantitative filter paper or through
a porcel~in filter crucible (this time·great care must be taken
to clean the beaker and stirrer from adhering precipitate),
and wash the precipitate with,cold.l : 2~ammonium hydroxide
solution until the washings give no turbidity with dilute nitric
acid and silver nitrate solution.
If a filter paper was used, fold it up, and place it in a pre-
viously igaited and weighed p,latinufll cruciple (1). Char the
paper slowly without allowing it to ignite, and burn off the
carbon at as Iowa temperature as possible with free access
of air (gradually increase the flame but do not heat the crucible
to more than the faintest red), and then ignite to constant
weight in an electric muffle furnace at 1050° to 1l00°C or,
less desirably, over a MC\k;r or Fisher high temperature burner.
If a porcelain filter crucible was used, dry it first in an air
oven at 100° to HO°C and then heat it gradually to 1050°
to 1100°C in an electric muffle furnace; if an electric furnace
is not available, place the porcelain crucible inside a nickel
crucible (Fig. 2-33), and heat it gradually to th:e full heat of a
Meker or equivalent burner. Heat for 25 to 30 minute
periods until constant weight is atta.ined.
Gravimetric A nalysi.~ 41}7
Calcu!ate the percentage of mag-ncsium in the compound.

The factor for converting Mg2~l)7 to Mg is 0·21851.
M.ote. 1. For erementary students it is sufficient to dry the filter with the preci-
pitate in the steam oven (or at lOooq. and to incinerate the filter papc!r apart
from the precipitate (Section D, 36C) at as Iowa temperature as possible; the
paper should not be allowed to take fire. After the volatile carbonaceous
matter has been -burnt off. the residue may be ignited strongly with the lid
of the crucible displaced. to allow circulation of the air, until the residue-is as'
white as possible. The main precipitate is added. and the whole ignited to
constant weight as described above.
For other methods for the determination of magnesium, see
Section IV, 46.
IV, 12. Determination of nickel as the dimethylglyoxime
complex.-Discussion. The nickel is precipitated by the
addition of an alcoholic solution of dimethyglyoxime
{CH3'C(: NOH).C(: NOH).CH 3 , referred to in what follows as
H.Dmg} to a hot- slightly add solution of the nickel compound,
and then adding a slight e~cess of ammonia solution. The
precipitate is washed and weighed as nickel dimethylglyoxime
after drying at 100° to 120°C (compare Section I, 62A) :
NiSO, + 2H.Dmg = Ni(Dmg)e + H 2S0 4 ,
The' precipitate is soluble in free mineral acids (even as
much as is liberated by reaction in neutral solution), in alco-
holic solutions containing more than 50 per cent by volume,
and in concentrated ammoniacal solutions of cobalt, but is
insoluble in dilute ammonia solution, in solutions of ammo-
nium .salts, and in dilute acetic acid-sodium acetate solutions.
Excessive amounts of ammonium hydroxide, cobalt or copper
retard the precipitation. Better results are obtained in the
presence of cobalt, manganese or zinc by adding sodium
or .ammonium acetate to precipitate the complex (compare
Section IV, S6D); ferric iron must, however, be absent.
Dimethylglyoxime forms sparingly soluble salts with
palladium, platinum and ferrous iron, whilst gold is reduced.
These elements and indeed all the elements of the hydrogen
sulphide group should be absent. Ferric iron, aluminium,
and chromic chromium are rendered in~ctive by the addition
of a soluble tartrate or citrate, with which these elements form
complex ions.
Dimethyglyoxime is almost insoluble in water, and is
added in the form of a 1 per cent solution in rectified spirit or
absolute alcohol; 1 ini. of this solution is sufficient for the
precipitation of 0·0025 g. of nickel. As already pointed ou't, .
precipitation is made in a hot feebly acid sollltion of a nickel
salt. and then rendering the solution faintly alkalin.e. This
procedure gives a more easily filterable precipitate than does
direct precipitation from cold'or from ammoniacal solut-ions.
Only a slight excess of the reagent should be used. since
dimethylglyoxime is not very soluble in water or in very
dilute alcohol and may precipitate; if a very large excess is
added (such that the alcohol content of toe solution exceeds
50 ,per cent), some of the precipitate may dissolve.
Procedure. A. NickeJ in a nickel salt. Weigh but accur-
ately 0·3 to 0·4 g. of pure (preferably A.R. *) nickel ammonium
sulphate NiSO,.(NH')2S0,.6H20 into a 400 ml. beaker pro-
v~ded with a clock glass cover and stirring rod. Dissolve it in
water. add 5 mI. of dilute hydrochloric acid (I : 1) and dilute
to 200 mI. Heat to 70° to 80°C, add a slight excess (25 to 30
mI.) of the dimethylglyoxime reagent, and il)lmediately add
dilute ammonia solution dropwise and ""ith constant stirring
until precipitation takes place and then in slight excess
(alternatively, 3 to 4 g. of sodium acetate may be used to
effect precipitationt). Allow to stand- on the s~eam bath for
20 to 30 minutes, and test the solution for complete precipi-
tation when the red precipitate has settled out. Di~est on
the water bath for 1 hour. Filter through a Gooch. sintered
glass or porous porcelain crucible, previously heated to 110° to
120°C and weigHed after cooling in a desiccator. Wash the
precipitate. with hot water until free from chloride, and dry
it at 110° to 120°C for 45 to 60 minutes. Allow to cool in a
desiccator and weigh. Repeat the drying until constant
weight is attained. Weigh as Ni(C 4H,02N2h. which contains
20·31 per cenf of Ni.
Calculate the percentage of nickel in the salt.
B. Nickel'in nickel steel. Weigh out accurately about 1 g.
of the drillings or borings of the nickel steelt (or sufficient of
the sample to contain 0·03 to 0·04 g. of nickel) into a 100 to '
150 ml. beaker or porcelain basin, dissolve it in the minimum
volume of concentrated hydrochloric acid (about 20 ml.
• This is obtainable from the Mallinckrodt Chemical WorkS. U.S.A.; see
Section A 9. Alternatively, sufficientofa nickel salt to contain about 0·05 g. of
nickel may be used.
t The sodilJm acetate method cannot be used in the presence of ferric iron;
:ferrous iron does not interfere.
! Ridsdale's .. Nickel Steel. No. 22 .. (one of the Analysed Samples for
Students) or the Bureau of Analysed Samples" Nickel Steel' T2 ' .. (a British
• Chemical Standard) is suitable. Both steels contain about 4 per cent of'
nickel.
should suffice}, and boil with successive additions of concen-
trated nitric acid (ca. 5 m!.) to ensure complete oxidation of
the iron to the ferric state. Dilute somewhat, filter, if neces-
sary, from any solid material, and wash the paper with hot
water; .dilute the filtrate (or solution) to 250 ml. in a 400 ml.
beaker. Add 5 g. of citric or tartaric acid, neutralise the
solution with dilute ammonium hydroxide solution, * and
then barely acidify «(litmus) with dilute hydrochloric acid.
Warm the solution to 60° to 80°C, add a slight excess of a 1
per cent alcoholic solution of dimethylglyoxime (25 to 30 mI.),
immediately followed by dilute ammonia solution until the
liR.uid is slightly ammoniacal, stir well, and allow to stand on
the steam bath for 30 to 60 minutes. Filter off the precipitate
through a weighed filter crucible; test the filtrate for complete
precipitation with a little dimethyl-glyoxime solution, and
wash the precipitate with hot water until free from chloride.(l)
Dry the precipitate at, HO° to l20°C for 45 to 60 minutes,
and weigh as Ni(C,H7 0 2N z)2'
Calculate the percentage of nickel in the steel.
Note. 1. A rapid method for weighing the precipitate: collected in a
sintered glass or porous porcelain crucible, is as follows (J. Dick, 1929).
After washing the nickel dimethylglyoxime with cold water, wash 5 or 6
times with rectified spirit, followed by several times with small volumes of
anhydrous ether. Suck the precipitate dcy at the pump for 10 minutes,
wipe the outside of the crucible with a clean linen cloth, and leave in a vacuum
desiccator for 10 minutes. Weigh as Ni(C,H70.N I ) •••
.
For other methods for the determination· of nickel, see
Section IV,_32.
"
• If a precipitate appears or if the solution is not clear when it is rendered
ammoniacal. more tartaric or citric acid must be added until a perfectly clear
.soluti9n is obtained upon adding dilute ammonia solutio!:..
SYSTEMATIC GRAVIMETRIC ANALYSIS
IV; 13. General discussion.-In the succeeding Sections a
brief account will be given of a number of selected methods for
the gravimetric determinations of the various elements and
raClicals. It is believed that these wiM suffice to meet the
needs of the student during the whole period of his training;
for a more detailed study, particularly of the limitations .of
some of the various methods, reference must be made to other
treatises. *
It has been considered e.xpedient, following Hillebrand and
Lundell, to treat the elements and radicals in the following
order:
(a) Elements that give precipitates with hydrogen sulphide
in acid solution: lead, silver, mercury, bismuth, cadmium,
copper, arsenic, antimony, tin, molybdenum, selenium, teIIu-
rimp, gold, palladium, platinpm (thallium). Of these the
sulphides of lead, silver, mercury, bismuth, cadmium, coppet
and palladium are insoluble in acids and in solu~ions <;>f alkali
sulphides. I
(b) Elements that give precipitates with hydrogen sulphide

in alkaline solution (or with ammonium sulphide) : ,alumi-
nium, beryllium, chromium, iron, nickel, cobalt, zinc,
manganese, vanadium, uranium, thorium, cerIum, titanium,
zirconium, thallium. Of these, iron, nickel, cobalt, zinc,
manganese, vanadium, uranitim 'and thallium form sulphides
that are soluble in acids, whilst the following elements form
hydroxides or basic compounds: aluminium, chromium,'
beryllium, thorium, cerium, zirconium and titanium.
(c) The alkaline earth elements (calcium, stron.tium and
barium) and, magnesium.
(d) The alkali elements: sodium, potassium and lithium
(ammonium). .
(e) Tungsten.
(f) The anions in the empirical order: chloride, bromide,
iodide, thiocyanate, cyanide, fluoride, chlorate, perchlorate,
• For example: W. F. Hillebrand and G. E. F. Lundell, Applied Inorganic
Analysis (J. Wiley, 1929); G. E. F. Lundell and]. 1. Hoffman, Outlines of
Methods of Chemical Analysis (J. Wiley, 1938); J. W. Mellor and H. V.
Thompson, A Treatise on Quantitative Inorganic Attaly~is (C. Griffin, 1938) ;
G. E. F. Lundell, J. I. Hofmann, and H. A. Bright, Chemical Analysis of Iron
and Steel (J. Wiley, 1931); W. W. Scott-No H. Furman, Standard Methods of
Chemical AnaZ',Isis, Vol. I (D. Van Nostrand Co.; The Technical ,Press, 1939).
500
iodate, sulphate, sulphide, sulphite, thiosulphate, phosphate,
phosphite, hypophosphite, oxalate, borate, silicate, fluosili-
cate, ferrocyanide, ferricyanide, nitrite, nitrate and carbonate.
IV,14. LEAD
DisCttssioll. Lead may be estimated in the following forms:
A. Lead sulphate, PbS04• This provides a separation from the.
·numerous elements which form soluble sulphates. Calcium,
strontium, barium, mercurY,.tungsten, silver, bismuth, antimony and
silica all interfere to a greater or less extent. Hydrochloric and nitric
acids exert a solvent action upon lead sulphate, hence if these are
present, the solution must be evaporated twice with sulphuric acid
'Until dense white fmnes are evolved. Lead sulpha.te is slightly
soluble in dilute sulphuric acid, although with the extremely dilute
acid, the solubility is l('ss than in water (0·0040 g. per litre at 18°C),
owing to the common ion effect; it is almost insoluble in alcohol.
.A satisfactory wash solution is a dilute solution of sulphuric .acid
that has been saturated with lead sulphate at the laboratory tem-
perature.
At a red heat, it is easily reduced by carbonaceous matter, with
loss of lead by volatilisation.· For this reason the precipitate is best
collected in a filtering crucible, and dried at 500 0 to 600°C. .
B. Lead chromate, PbCr0 4• This determination is more accurate
than A. but is limited in its applicability because of the general
insolubility of chromates.
C. Lead molybdate, PbMo04 • This is an excellent method since
the substance has a high molecular weight, is less solubl!-= than the
sulphate, and suffers no change upon ignition. Substances which
form insoluble molybdates (e.g., the alkaline earth metals and cad-
mi).lm), which are easily hydrolysed (~.g., tin or titanium), and which
form insoluble compounds with lead (e.g .. chromates, arsenates or
phosphates) must be absent.
D. Lead dioxide, Pb0 2 • This excellent electrol¥tic method is
described in Section IV, 79.
4. Determination of lead as sulphate. Proc$dure. Weigh
out accurately about 0·3 g. of the lead salt* into a 10 cm.
porcelain basin, dissolve it in abbu't 20 ml. of water, add 2 ml,
of pure concentrated sulphuric acid, evaporate the mixture as
far as possible on the water bath, then on a sand bath, wire
gauze or low temperature hot plate until thick white fumes
of sulphuric acid are freely evolved. This ensures the com-
plete removal of nitric acid and .conversion of the lead salt
into the sulphate.t Cool, dilute carefully with 40 ml. of
• A.R. lead nitrate is suitable. If the substance is insoluble in water, dis-
solve.it in the minimum quantity of nitric acid. I
t For work demanding the highest accuracy, allow to cool, ·wash down the
sides of the dish. and repeat the evaporation;
water (the final solution should contain about 4 mI. of con-
centrated sulphuric acid in 100 mI. of solution), mix thor-
oughly, and allow to stand for at least an hour. Filter off
the precipitate of lead sulphate through a weighed porcelain
or silica Gooch crucible or through a porous porcelain filtering
crucible. Wash the precipitate three times with 3 per cent
'(by volume) sulphuric acid saturated with lead sulphat~
(compare Section IV, 57A). Dry the crucible and contents
at 100°C for 1 hour, and then place it inside a nickel crucible
(Fig. 2-33) and heat gradually to a dull red heat (500° to 60Q°C)
until constant weight is obtained. Alternatively, the preci-
pitate, collected on a sintered glass or porous porcelain crucible
and washed as. above, may be washed with rectified spirit
until free from sulphate (1), and then dried at 130°C to con-
stant ~eight. Weigh as PbS0 4•
Calculate the percentag~ of lead in the sample. Weigh
as PbSO,.* I •
Note. 1. For the rapid weighing of the lead ~ulphate, the alcohol washing
is followed by several washings with small volumes of anhydrous ether, and
-the precipitate is sucked dry at the pump for 10 minutes. The outside of the
crucible is then wiped with a clean linen cloth, left in a vacnum desiccator
for 10 minutes, and weighed (}. Dick, 1929).
B: Deterrilination of lead as chromate. Procedure (compare
Section m, 109). Weigh out 0·3 g. of the lead salt (1), dissolve
it in 150 mI. of water, and add dilute acetic acid to the solution
until it -is distinctly acid. Heat to boiling, add from a
pipette 4 per cent potassium chromate solution in slight
excess (10 ml. will usually suffice). Boil gently for 5 to 10
minutes (or until the precipitate settles); the supernatant
liquid must be coloured slightly yellow. Filter through a
Gooch, sinter~d glass or porouS' porcelain filtering crucible,
wash' thoroughly with a hot dilute soh,ltion of sodium acetate
or with hot water. Dry at 120°C to constant weight. Weigh
as PbCr0 4 •
Note. 1. A.R. lead nitrate is suitable for practice in this estimation. If the
solution is neutral or basic, add acetic acid until it is distinctly acid. If the
solution contains nitric acid, add sufficient sodium acetate (5 to 10 g.) to replace
the nitric acid by acetic acid. '
C. Determination of lead as molybdate. Procedure. Weigh
out 0·3 g. of the lead salt, dissolve it in 200 ml. of water, and
add 4 drops of concentrated nitric acid. Heat to boiling, and
slowly add from a burette or pipette, with stirring, a 2·5 p~r
cCfnt aqueous solution of ammonium molybdate. When
• The conversion factors for all the determinations described in this and the
following Sections are collected in the Appendix (Section A. 2).
.
precIpitation
Gravimetric Analysis
~ppears to be complete, boil'for I minute; allow
503
the precipitate to settle, and add a few drop~ of the precipi-

tant to the supernatant liquid. If a precipitate forms, repeat
the process until 'the ammonium molybdate is present in
slight excess. When precipitation is complete, add dilute
ammonia solution (1 : 2) dropwise until the solution is neutral
or slightly alkaline to litmus. Acidify with a few drops of
acetic acid, and allow to stand for a few minutes. Decant
the supernatant liquid through a weighed porcelain or silica
Gooch crucible (or through a porcelain filter crucible), and
wash the precipitate 3 or 4 times by decanta-;-ion with 75 ml.
portions of 2 per cent ammonium nitrate solution. Transfer
the precipitate to the filter, and wash until the washings giv.e
no test for molybdenum (e.g., no brown precipitate with
potassium ferrocyanide. solution). Place the filter ·crucible
inside a nickel crucible or in a radiator(Fig. 2-14),and gradually
heat to dull redness. Maintain the crucible at dull redness
for 10 minutes, cool in a desiccator and weigh. -Repeat
the heating, etc., until constant weight is attailled. Weigh as
PbMoO,.
IV, 15. SILVER

Discussion. This element is usually-determined as silver chloride,
AgCl. The theory of the-process has been given under Chlorid'e
(Section IV. 5). Lead, cuprous, palladous, mercurous.and thallous
ions interfere as do cyanides and thiosulphates. If a mercurous (or
cuprous or thaUous) salt is present, it must be oxidised with concen-
trated nitric acid before the ,precipitation of silver; this process also
destroys cyanides and thiosulphates. If lead ffi present, the solution
must be diluted so that it contains not more than 0·25 g. of the
substance in 200 mI., and the hydrochloric acid must be added very
slowly.
Silver may also be determined electrolyticillly; for details see
Section IV, 83.
Determination of silver as chloride. Procedure.' The
solution (200 ml.) should contain about 0·1 g. of silver (I)
and about 1 per cent by volume of nitric acid. Heat to about
70°C; and add approximately O'2N pure hydrochloric acid
slowly and with constant stirring until no further precipitation
OCCUl;S; avoid a large excess of the acid. Warm until the
precipitate settles, allow to cool to about 25°C, and test the
supernatant liquid with a few drops of the acid to be sure
that pr.ecipitation is complete. Allow the precipitate to settle
in a dark place for several hours or, preferably, overnight.
Po_ur the supernatant liquid through a weighed Gooch of
sintered glass or porous porcelain crucible, wash the precipitate
by decantation with O'IN nitric acid, transfer the precipitate
to the crucible, and wash again with O'OlN nitric acid until
free from chloride. Dry the prc~ipitate first at lOO°C and
then at 130° to 150°C, allow to cool in a desiccator and weigh.
Repeat the heating, etc., until constant weight is obtained (2).
Weigh as AgCl.
Notes. 1. For example. from 0·2 g. of A.R. silver nitrate.
2. Sec footnote in Section IV, 5.
IV,16. MERCURY
-Discussion. Mercury may be estimated in th(' following forms:
A. Meljiluric sulphide, HgS. The precipitation of mercury in
hydrochloric acid solution by hydrogen sulphide as mercuric sul-
phide is an accurate procedure in the absence of copper, cadmium,
tin, zinc and thallium; the latter metals complicate reactions which
are based upon the behaviour of pure mercuric sulphide. Unless
the experimental.conditions detailed below arc strictly followed, the
precipitate is liable to be contaminated with a little sulphur, which
must be removed by extraction with carbon disulphide. Oxidising
agents (nitric acid, chlorine, ferric chbride, etc.) must be absent.
B. Mercury zinc thiocyana.te, Hg[Zn(CNS)4]' This method is
based upon the precipitation of mercury from mercuric compounds
iI1' neutral or acid solutions as mercurie zinc thiocyanate with a
reagent whicq contains 39 g. of ammonium thiocyanate and 29 g.
of zinc sulphate per litre. For quantitative precipitation not morc
than 5 per cent of free acid should be present in the solution before
the addition of the precipitating agent. If a larger quantity of acid
is present, it must b~'neutralised with sodium hydroxide solution:
ammonia solution should ndt be used because an excessive quantity
of ammonium salts exerts a solvent action upon the precipitate.
Cadmium, cobalt, copper, bismuth, manganese and mercurous
compounds must be absent since they ~ive insoluble thiocyanates
with the reagent; small quantities of nickel do not interfere:
C. Copper ethylenediamine mercuri-iodide [Cu en~JLHgI.J; ethy-
lenediamine method. Th~s is a rapid method for the determination
of mercury.
For further discussion, see Section I, 62P.
A. Determination of mercury as sulphide. Procedure.
Weigh out accurately about 0·15 g. of the mercuric salt (1),
dissolve it in 100 ml. of. water, and add a few ml. of dilute
hydrochloric acid. Saturate the cold solution with washed
hydrogen sulphide (2), allow the precipitate to settle, and filter
thr?ugh a weighed Gooch, sintered glass or porous porcelain
crucible. Wash the precipitatE' with cold water (:l), and
WE'igll it, as HgS, after drying at 105° 'to ] lOoe (4).
Notes. 1. A. R. mercuric chloride is suilahle. Alternativelv. the solution
should contain not more than 0'1 g. of mercuric mercury per 100 mI., and
should be free from oxidising agents.
2. The colour of the mercuric sulphide precipitate will become perfectly
black as soon as the liquid is saturated with the gas.
3. If the presence of sulphur is suspected, the precipitate is washed with hot
water, alcohol, carbon bisulphide, or alcohol + ether, and then dried at 100°
to llOce. •
4. A rapid method for weighing the precipitate, collected in a sintered
glass or porous porcelain crucible, is as follows U. Dick, 1929). After washing
the mercuric sulphide with cold water, wash it ;") or 6 times with rectified spirit,
followed by several times with small volumes of anhydrous ether, and then
suck the precipitate dry on the pump for 10 minutes. Wipe the outside of the
crucible dry with a clean linen cloth and le~lV(.' in a vacuum desiccator for
10 minutes. Weigh as HgS,
B. Determination of mercury ,as mercuric zinc thiocyanate.

Procedure. Weigh out accurately about 0,2 g, of the mercuric
salt (e.g" A,R. mercuric chloride) into a 250 ml. beaker, and
dissolve it in 100 ml. of cold water containing 2 tlrQps !Jf
concentrated hydrochloric acid. Add 25 mL of the ammo-
nium thiocyanate-zinc sulphate reagent (see DisCltSsion).
Stir vigorously with a previously moistened glass rod for 2 to
3 minutes, but avoid. touching the walls of the beaker. The
initially clear liquid gradually becomes turbid during the
stirring, and a white crystalline precipitate slowly forms.
Allow the contents of the beaker to stand for at least 2 hours.
Filter the precipitate on a weighed sintered glass or porous
porcelain crucible; transfer the precipitate to the crucible
with the aid of a wash-liquid composed of 5 ml. of the precipi-
tating reagent and 450 ml. of water. (The precipitate is
slightly soluble in water, but practically insoluble in the special
wash liquid.) Finally, wash the precipitate [) times with the
special wash-liquid, and then once with a little water to
remove the adhering wash-liquid. Dry the cnicible and
precipitate .in an electric oven maintained at 105° ± 2°C
untif constant in weight. Weigh as Hg[Zn(CNS)4].
C. Determination of mercury by tlJe ethylenediamine
method. Procedure. The neutral or faintly ammoniacal
mercury solution should occupy a volume of 80 to 500 ml.,
and contain not more than about 0·2 g. Hg. Treat with
excess (say, 2-3 g.) of K1 in the form of a 2% solution, heat
nearly to boiling and precipitate with a boiling, concen-
trated aqueous solution of copper ethylenediamine nitrate
[Cn en 2](N0 3 )2 or of the corresponding sulphate (Section
I, 62P). Allow to cool to room temperature and filter off
the dark blue-violet, crystalline complex [Cu en 2J[HgI 4J on a
weighed crucible (Gooch, sintered glass or porous porcelain).
Wash 3 to 6 times with a solution containing 0·2 g. of KI
and 0·2 g. of (Cu en 2](NO a)2,2H2,'O (or the corresponding
sulphate) in 200 m!. of water. To remove the water adhering
to the wall of the crucible, wash 3 or 4 times with 2 ml.
portions of 95 per cent alcohol and finally 3 or 4 -times with 2
ml. portions of anhydl;ous ether. Suck the precipitate dry
at the pump for 10 minutes, and dry the precipitate for 10
minutes (or to constant weight) in a vacuum desiccator.
Weigh as [Cu enJ[HgIJ. .
For the determination of mercury in aqua regia solution (up
to 0·3 g. Hg in 100 to. 200 ml.),'render the solution f!iintly
alkaline with ammonia solution, add a large excess of potas-
sium iodide.(say, 3 to 4 g. KF in the form of a 2 per cent solu-
tion), heat to boiling and precipitate as above. When pre-
cipitation is complete, dilute with one-quarter of the total
volume with distilled water, allow to cool ~lowly to room
temperature, and complete the determination as already
detailed.
IV, 17. BISMt1TH
Discussion. Bismuth may be satisfactorily determined in the
following forms:
A. Bismuth oxyiodide,. BiOI. The cold bismuth solution, weakly
acid with nitric acid, is treated with an excess of potassium iodide
when BI3 and some K[Bil,] are formed: ...
Bi(N0 3)3 + 3KI = BiI3 + 3KNO a ;
Bl3 + Kl = K[BiI4]. '
Upon dilu,tion and boiling, bismuth oxyiodide is formed, and is
weighed as such after suitable drying.
BIa + H 20 = BiOI + 2HI;
(black) •
KlBiIJ + H 20 = BiOI + KI + 2HI.
(yellow)
A large excess 'of potclssium iodide should be avoided since the com-
plex salt is not so readily hydrolysed as the tri-iodide. This is an
excellent method, because the oxyiodide is precipitated in a form
which is very convenient for filtration and weighing.
B. Bismuth pyrogallate, Bi(CoH30a)' The precipitation of bis-
muth with pyrogallol is quantitative only if the acidity (hydro-
chloric, sulphuric or nitric acid) does not exceed O·IN. The method
is an excellent one for the determination of bismuth in the presence
of lead.
C. Bismuth oxide, Biz0 3, through the cupferron complex. The
bismuth in hydrochlori<; or njtric acid solution is precipitated as the
cujJferron complex (see Section I, 62B), and the latter is converted
by ignition into, and weighed as, Bi z0 3. This procedure separates
'bismuth from cadmium, lead, arsenic, mercury, antimonic and zinc
ions. •
. D. Bismuth" oxinate," Bi(C9~ON)3' This method uses oxine as
the precipitating agent. For the advantages and limitations of
this versatile reagent, see Section I, 62C. Halogens must be
absent as there is a tendency f~r compounds of the general formula
(C 9H 70N)HBiX, (where X = halogen) to be precipitated.
A. Determination of bismuth as oxyiodide. Procedure.
The cold bismuth nitrate solution, containing 0·1 to 0'l5 g.
of Bi (1), must be slightly acid wit!). niti:ic acid (2), and occupy
a volume of about 20 ml. Add finely-powdered solid potas-
sium iodide, slowly and with stirring, until the supernatant
liquid above the black precipitate of bismuth tri-iodide is just
coloured yellow (due to K[BiIJ). Dilute to 200 Illl. with
boiling water, and boil for a few minutes. The black tri-
iodide is converted into the copper-coloured precipitate of the
oxyiodide. The supernatant liquid should be colourless;
if this is yellow, a further 100 ml. of water should be added,
and the boiling continued until colourless. Add a few drops
of methyl orange indicator, and then sodium acetate solution
(25 g. per litre) from a burette until the solution is neutral.
Filter off the precipitate through a weighed Gooch, sintered
glass or porous porcelain crucible, wash with hot water, .and
dry at 105 to 1l0°C to constant weight. Weigh as BiOI.
0
Notes. 1. A suitable solution for practice can. be obtained by dissolving

about 0·15 g. of pure bismuth (the A.R. product is obtainable from the Mallinc-
krodt Chemical Works, U.S.A.; see Section A, 9), accurately weighetl. in the
minimum volume of 1 : 4 nitric acid.
2. Chloride and bromide should be absent. If the solution is strongly acid
with nitric acid, it should be evaporated to dryness on the water bath, and
dissolved in a little dilute nitric acid.
B. Determination of bismuth as pyrogaUate. ProcedtJre.
The solution (150 ml.) should be weakly a.cid with nitric acid
and contain 0-1 to 0·2 g. of Bi. Treat the solution with dilute
ammonia solution until a permanent turbidity is obtained;
render the solution clear by tHe cautious addition of a little
dilute nitric acid. Heat to boiling, and add a slight excess of
a solution of pure pyrogallol (Section 1,62.0) in air-free water.
A yellow, finely crystalline precipitate is immediately formed.
Boil for a short time, test for completeness of precipitation
with a little of there agent, dilute slightly, and filter through
508 Q11alltitath,c Illorgllllic A Ilalysis
a weighed sintered glass Of porous porcelain crucible after
the precipitate settles out. Wash with O'05N nitric ac~d,
and finally· with water. Diy to constant weight at 110°C.
Weigh as Bi(CGH303)'
C. Determination of bismuth as oxide after precipitation as
the cupferron complex. Procedure. Treat the aqueous bis-
muth nitrate solution (200 ml.) with freshly prepared 6 per
cent aqu~ous cupferron solution until there is present at least
three times of cupferron as of bismuth (i.e., 1:: 0·3 g. per O·lg.
Bi). Heat the solution to boiling, add ammonia solution
until the liquid is alkaline (litmus paper), followed by dilute
nitric acid u,ntil it is just acid. Filter the precipitate through
a quantitative filter paper, and wash with 0-1 per cent cup-
ferron solution containing a few drops of dilute nitric acid.
Ignite in a crucible in the usual manner (Section II, 36), and
finally with free access of air over the full flame of a Bunsen
burner to constant weight. Weigh as Bi 20 a.
D. Determination of bismuth with .. oxine." Procedure.
The solution of bismuth nitrate (100 ml.) should contain
sufficient tartaric acid to prevent precipitation in the follow-
ing stages. Add a drop or two of phenolphthalein, and then
dilute arpmonia solution until the indicator is just pink.
Acidify with acetic acid, and add 1 to 2 g. of ammonium or
sodium acetate for each 0·05 g. of Bi expected to be present.
Heat to 60° to 70°C, treat rapidly with a saturated alcoholic
solution of oxine (excess is without influence), then heat
almost to boiling, and allow to cool. Filter off on a weighed
sintered glass or porous porcelain crucible, wash with hot
water, and dry to constant weight at 130° to 140°C. Weigh as
Bi(C 9 H 60N )s.
IV, 18. - CADMIUM
Discussion. Cadmium may be determined in the following forms:
A. Cadmium molybdate, CdMo0 4 • The method is of limited
application, since all metals other than magnesium and the alkali
metals must be absent.
B. Cadmium .B-naphthaquinoline complex, [(CI3~N)2H2JCdI4'
This method permits of the separation of cadmium from relatively
large quantities of zinc, iron, chromium, aluminium, cqbalt, nickel,
manganese and magnesium. and also from antimony and tin if
ammonium oxalate or larger amounts of sodimll tartrate are u~eJ.
C. Cadmium quinaldinate, Cd(CIAO~)2' Quinaldinic acid or its
sodium- salt p;recipitates cadmium quantitatively from acetic acid
or neutral, solutions. The precipitate is collected on a Gooch type
of crucible, and dried at 125°C. A determination may be completed
in about !IO minutes. For the limitations of the method, see Section
I. 62N.
D. [Cd(CsHsN)J(CNS)2; pyridine method. If a hot neutral or
faintly acid solution of a cadmium salt is treated with ammonium
thiocyanate and pyridine, dipyridine cadmium thiocyanate is
quantitatively precipitated. This precipitate is collected, and
washed inter alia with alcohol and ether containing a little pyridine;
it may be dried simply by leaving in a vacuum desiccator for 15 to
20 minutes. A determination can thus be completed in less than an
hour. The limitations of the method are discussed under Copper
(Section IV. 19E). If the solution is weakly acid, ammonium thio-
cyanate may be added, followed by pyridine, until a precipitate just
forms, the latter dissolved by. warming, and a further 1 m!. of pyri-
dine added.
E. Metallic cadmium. This excellent electrolytic method is
described in Section IV, 80.
A. Determination of cadmium as molybdate. Proced1tre.
Weigh out accurately 0·4 to 0·5 g. of the cadmium salt (I),
dissolve it in 100 mI. of water, render the solution just alkaline
with ammonia solution, and then feebly acid with acetic
acid. Treat the boiling solution dropwise and with constant
stirring (2) with a solution of ammonium molybdate (3) in
slight excess (4), and maintain the mixture at about 90°C
until the precipitate has settled. Allow to stand for at least
2 hours, but preferably overnight. Filter off the precipitate
of cadmium molybdate upon a weighed sintered glass crucible
of fine porosity, or upon a Gooch crucible covered with a layer
of ground asbestos, wash with hot water, and dry at 120°C to
constant weight. Weigh as CdMoO,.
Notes. 1. A.R. cadmium iodide is a suitable salt to employ for practice in
this determination.
2. The precipitO\.te adheres very tenaciously to the glass if this is touched
with the stirring rod during precipitation. Touching of the sides of the glass
vetsel with the stirring rod should therefore be avoided.
3. This solution is prepared by stirring up 15 g. of A.R. molybdenum
trioxide with 200 ml. of water, adding a few drops of phenolphthalein, and then
concentrated ammonia solution with constant stirring until the oxide is
dissolved. Acetic acid is then added urltil the pink colour disappears, but the
reaction is acid to litmus: Filter and dilute to 500 ml. 1 !Ill. of this reagent
will precipitate 0·02 g. of Cd.
4. Add a drop of the supernatant solution to a saturated solution of pyro-
gallol in chloroform; a brown colouration is produced if excess of molybdate
is' present.
B. Determination .of cadmium as the ,B-naphthaquinoline
complex. Procedure. The cadmium salt solution, containing
about 0·15 g. of Cd, should occupy a volume of about 50 m!.
and be 2N with regard to sulphuric acid. Add 50 m!. of 10
per cent sodium tartrate solution,' followed successively by a
2·5 per cent solution of ,B-naphthaquinoline in 0·5N sulphuric
acid, a few drops of dilute sulphuric acid, and then 0·2N potas-
sium iodide in excess. After 20 minutes, filter the precipitate of
the cadmium complex through a weighed Gooch, sintered
glass or porous porcelain crucible, wash with a solution con-
taining 10 ml. of·0·2N potassium iodide, 10 ml. of 2·5 per cent
,8-naphthaquinoline in 0·5N sulphuric acid, 80 ml. of water and
1 to 2 drops of dilute .sulphurous acid, and finally suck as free
as possible from the wash liquor. Dry the precipitate to
constant weight at 130°C. Weigh as [(C13H9N)2H2](CdI,).
C. Determination of cadmium as quinaldinate. Procedure.
The solution (150 ml.) should be neutral or weakly acid with
acetic acid, and should contain 0·1 to 0 ·15 g. of Cd. Heat the
solution to boiling, and remove the source of heat. Add
the reagent (a 3·3 per cent solution of quinaldinic acid or of
the sodium salt in water) dropwise with vigorous stirring
until present in slight excess. Then neutra1ise carefully with
dilute ammonia solution, and allow the white curdy precipi-
tate to settle. When cold, wash with cold water by decan-
tation, filter through a sintered glass or porous porcelain
crucible, wash thoroughly with cold water, and dry at 125°C
to constant weight. Weigh as Cd(C 1oH 6 0 2 N)2'
D. Determination of cadmium by the pyridine method.
Procedure. The solution (75 to 100 m!.) should contain about
0·1 g. of Cd (1) and be neutral or very feebly acid. Add 0·5
to 1·0 g. of A.R. ammonium thiocyanate, stir, heat to boiling,
and treat the solution with 1 ml. of pure pyridine dropwise
and with stirring. The,complex slowly separates as the
solution cools. Filter the cold solution through a weighed
sintered glass or porous porcelain crucible, transfer the pre-
cipitate to the crucible with the aid of solution 1. Wash 4 to
5 times with solution 2, then twice with 1 ml. portions of
solution 3, ,and finally 5 to 6 times with small volumes (ca. 1 ml.)
of solution 4. (For further experimental details, see under
Copper, Section IV. 19E, and Zinc, Section IV. 34D.) Dry
the precipitate in a vacuum desiccator (Fig. 2-10) for 10 to 15
minutes ~nd weigh. Repeat the drying until constant
weight is attained. Weigh as [Cd(CsHsN)2J(CNSh.
Solution 1. 100 Ml. of water containing 0·3 g. of NH 4CNS and
0·5 m!. of pyridine.
Solution 2. 73 MI. of water, 25 ml. of 95 per cent alcohol, 0·1 g. of
NH 4CNS, and 2 ml. of pyridine.
Solution 3. 10 MI.·of absolute alcohol and 1 ml. of pyridine.
Solution 4. 15 MI. of absolute (sodium dried) ether and 2 drops of
pyridine.
Hote. 1. For practice in this determination the student may employ about
o3 g .• accurately weighed. of A.R. cadmium sulphate or of A.R. cadmium
iodide.
IV,19. COPPER
Discussion. Copper may be determined in the following forms:
A. Cuprous thiocyanate, Cu2 (CNS)2' This is an excellent method
since most thiocyanates of the metals are soluble. Separation may
thus be effected from bismuth, cadmium, arsenic, antimony, tin,
iron, nickel, cobalt, manganese and zinc. The addition of 2 to 3 g.
of tartaric acid is desirable for the prevention of hydrolysis when
bismutl1, antimony or tin are present. Excessive amounts of
ammonium salts or of the thiocyanate precipitant should be absent
as should also oxidising agents and high acidity. Lead, mercury,
the precious metals, selenium and tellurium interfere and contami-
nate the precipitate.
B. Copper benzoinoxime, Cu(C14H1l0~). a-Benzoinoxirne (cup-
ron) is a specific reagent for the determination of copper in ammo-
niacal solutions (compare Section I. 62D). A green, heavy and
readily filterable precipitate is obtained: this is insoluble in water,
dilute ammonia solution, acetic acid, tartaric acid and alcohol, is
slightly soluble in concentrated. ammonia solution, and readily
soluble in mineral acids. Precipitation is quantitative in ammonia-
cal tartrate solutions: separation can thus be effected from iron
and other metals whose hydroxides are not precipitated in tartrate
solutions. Separation can also be made from cadmium. zinc, cobalt
and nickel, which are not precipitated in'ammoniacal solutions.
C. Copper salicylaldoxime, Cu(C7H50~)2' In this method silver,
mercuric, arsenious, ferric, nickel, cobalt and zinc ions do not
interfere provided. the solution contains sufficient acetic acid; the
copper complex alone is insoluble in acetic acid. In neutral or in
very faintly acid soiution nickel and cobalt are precipitated, but the
ot~er ions remain in solution. An excellent method for the separa-
tion o£'copper ,and nickel is thus available (see Section IV, 32B).
D. Copper quinaldinate. Cu(C 1oHaON 2)2,H zO. Quinaldinic acid or
its sodium salt 1S a specific reagent for copper if precipitation is
made in sulphuric acid solution and iron and zinc are absent (com-
pare Section I. 62N). With this reagent, a determination of copper
may be completed in 2 hours. -
E. [Cu(CsHsN)J (CNS) 2 ; pyridine method. If a neutral or faintly
acid copper solution is treated with ammonium thiocyanate and
pyridine, dipyridine copper dithiocyanate is quantitatively 'pre-
cipitated. This precipitate is collected and washed inter alia with
alcohol and ether containing a little pyridine, and then dried simply
by leaving in a vacuum desiccator for 10 to 20 minutes. A deter-
mination can be completed in less than 1 hour·. This rapid method
is applicable in the absence of cadmium, nickel, cobalt, zinc and
manganese (sec Section I, 62L). Large quantities of ammonium
salts must be absent as these exert a solvent action upon the preci-
pitate. If the solution is strongly acid, it must be evaporated to
dryness and the residue dissolved in water.
F. Copper ethylenediamine mercuri-iodide [Cu en 21[HgI41; ethy-
lenediamine method. This, like E, is a rapid method for the deter-
mination of copper. Owing to the small conversion factor (O'0712())
the procedure is v.aluable for the determination of small amounts of
copper.
For further discussion, see Section I, 62P.
G. Metallic copper. The electrolytic method is one of the best
for the determination of copper. Comparatively large quantities
(up' to 5 g.) can be handled (see Section IV, 78 for details).
A. DeterIIli.Ii."ation of copper as cuprous thiocyanate. Pro-
cedure. Weigh out accurately about 0·4 g. of the copper salt
(1) into a 250 ml. beaker, and dissolve it in--50 ml. of water.
Add a few drops of dilute hydrochloric acid, and then,a slight
excess (2 to 3 ml.) of freshly-prepared saturated sulphurous
acid solutioq. Dilute the c,old liquid to about 150 ml., heat to
boiling, and add freshly-prepared 10 per cent ammonium
thiocyanate solution, slowly and with constant stirring, from
a burette until present in slight excess. The precipitate of
cuprous thiocyanate should be white; the mother liquor
should be colourless 'and smell of sulphur dioxide. Allow to
stand for several hours, but J)referably overnight. Filter
through a weighed filter crucible (Gooch, sintered glass or
porous porcelain), and wash the precipitate 10 to 15 times
with a cold solution prepared by adding to every 100 m!. of
water 1 ml. of a 10 per cent solution of ammonium thiocyanate
and 5 to 6 drops of saturated sulphurous acid solution, and
finally several times with 20. per cent alcohol to remOve
ammonium thiocyanate (2). Dry the precipitate to constant
weight at 110° to 120°C (~). Weigh as CUa(C~Sh.
Notes. 1. A.I\. copper sulphate pentahydrate is suitable for practice in this
determination. 0·4 Gram of this contains about 0·1 g. of Cu.
:!. Alternatively, but less desirably, the<precipitate may be washed with cold
water until the filtrate gives only a slight reddish colouration with ferric
chloride, and finally with :!O per cent alcohol.
3. The precipit..tl', collected in a sintcred glass (lr porous porcelain crucible,
may be weighed more rapidly as follows. (J. Dick, 1930). Wash the cuprous
thiocyanate 5 or 6 times with rectified spirit, followed by a similar treatment
with small volumes of anhydrous ether, then suck the precipitate dry at the
pump for 10 minutes, wipe the outside of the crucible Wlth a clean linen cloth
and leave in a vacuum desiccator for 10 ~inutes. Weigh as Cu 2 (CNS)z.
B. Determination of copper with a-benzoinoxime. Pro-

cedure. Treat the neutral solution, which should be free
from ammonium salts and contain not more than 0·05 g. of
copper, with dilute ammonia solution until a clear blue solu-
tion is obtained. Heat to boiling and precipitat~ the copper
by the addition, dropwise, of a 2 per cent alcoholic solution
of t~e reagent. Precipitation is complete when the blue
colour of the solution disappears. Filter the heavy green
precipitate on a weighed sintered glass or porous porcelain
crucible; wash with hot 1 per cent ammonia solution, then
with warm alcohol, and finally with hot water. Dry to con-
stant weight at 105° to 115°C. Weigh as Cu(C 14H l1 0 2N). It
is recommended that completeness of washing be tested for by
washing the dry precipitate with warm alcohol followed by
-hot water. Dry again to constant weight at 105° to lI5°C.
C. Determination of copper with salicylaldoxime. Proced1tre.
Add 2N sodium hydroxide to the copper solution (100 ml.),
containing 0·05 to 0·1 g. of Cu, until a slight permanent pre-
cipitate is for~ed, and dissolve this by the addition of a little
-dilute acetic acid.'" Add the salicylaldoxime reagent (for
preparation, see Section I, 62E) in slight excess at room tem-
perature. Filter off the precipitated complex on a weighed
sintered glass or porous porcelain crucible, wash with water
until the washings give no colour with ferric chloride, and dry
to constant weight at 100° to 105°C (about 1 hour). Weigh
as Cu(C 7 H s0 2N)2' Decomposition occurs above 115°C.
D. Determination of copper as quinaldinate. Procedure.
Acidify the solution (ca. 150 ml.), containing up to q·l g. of
copper, with 2 to 5 m!. of 2N sulphuric acid, heat to boiling,
and add the reagent (for preparation, see Section I, 62N)
slowly and with stirring until present in slight ex€ess. The
green crystalline precipitate of copper quinaldinate separates
out immediately. Wash with hot water by decantation, then
transfer to a weighed sinTered glass or porous porcelain crucible
and wash with hot water until the excess of the reagent
is removed. Dry at 125°C to constant weight. Weigh as
Cu(C 1o H 6 0:!,N)2,H 2 0.
E. Determination of copper by the' pyridine method. .Pro-
cedure. The solution (100 ml.) should contain not more than
* Quantitative precipitation of copper commences at pH 2·6.
s
0·1 g. of Cu (1) and be neutral or very faintly acid. Add
pure ,pyridine dropwise until the colour of the solution is
practically azure-blue (1 to 2 ml. should suffice), then add 0·5
g. of. solid. A.'R. ammonium thiocyanate, and stir vigqrously.
After a few minutes, collect the precipitate on a weighed
sintered glass or porous porcelain crucible; transfer the
precipitate to the crucible with the aid of solution 1. Wash
the precipitate 6 to 8 times with solution 2, then wash the
walls of the crucible 2 to 3 times with 1·5 ml. portions of
solution 3 (use a 2 rol. pipette for this purpose), and finally
several times with small volumes (1 to 2 ml.) of solution 4. It
is important to suck the precipitate almost dry. beh,:een each
washing; it is advisable also to stir the precipitate with a
thin glass rod during the washing with solutions 3 and 4.
Dry the crucible 'and precipitate in a vacuum desiccator (Fig.
2-10) for 5 to 2Q minutes, according to its weight. Repeat
the drying until constant weight is attained. Weigh as
[Cu(C.H.N)2J(CNS)2· -
Solution 1. 100 Ml. of water containing 0·3 g. of NH 4CNS and 0·3
m!. of pyridine.
Solution 2. 80 Ml. of 95 per cent alcohol, 19·2 m!. of water, 0·8
m!. of pyridine, and 0·05 g. of NH4 CNS.
Solution 3. 10 MI. of absolute alcohol + 2 drops of pyridine.
Sol1ttion 4. 20 Ml. of absolute (sodium dried) ether + 2 drops of
pyridine.
Note. 1. For practice in this determination, the student may, employ ca.
g., accuz:ately weighed, of·A.R. copper sulphate.
()·2li
F. Determination of copper by the ethylenediamine method.

Procedure. The solution should be neutral or very faintly
acid and should contain up to 0·10 g. of Cu in a volume of 100
to 1000 ml. The experimental details for a volume of 100 ml.
will be -given. Treat the solution dropwise w(th an aqueous
solution of ethylenediamine until the tharacteristic coloura-
tion of th.e [Cu en 2J++ ion appears; add a little more of the
reagent, followed by 1 to 2 g. of solid ammonium nitr~te and
2 g. of potassium iodide. Heat to boiling, add. a hot concen-
trated sol\ltion of potassium mercurttiodide (1) to precipitate
the complex. Allow to cool slowly with frequent stirring of
the soluti9n. When cold, filter off the precipitated complex
[Cu en 2J[HgIJ through a weighed crucible (Gooch, sintered
glass or porous porcelain); transfer the -precipitate to the
crucible with the aid of a wash-liquid containing 0·1 g. of
mercuric chloride, 2 g. of potassium iodide, 1 g. of ammonium
nitrate and 2 to 3 drops of ethylenediamine per 100 mI., and
wash it severat 'imes with the solution. Wash -the walls of
the crucible 4 times with 2 ml. portions'of 95 per cent alcohol
(rectified spirit), followed by 3 to 4 times with 2 ml. portions
of anhydrous ether, and then suck the precipitate dry at the
pump for 10·minutes. Dry the outside of the crucible With a
clean linen cloth, l~ve it in a vacuum desiccator for 10
.minutes, and weigh. Repeat the final drying in the desiccator
until the weight is constant. Weigh.as [Cu en 2J[HgI 4).
Note. 1. The potassium mercuri-iodide solution is prepared by treating
aqueous mercuric chloride with a slight excess of potassium iodide solution.
,
lV.20. ARSENIC
Discussion. Arsenic may be determined in the following forms:
A. Arsenic trisulphide, AszS a• The arsenic must be present in the
tervalent state. Arsenic in the tervalent state (ensured by the
addition of, for example, ferrous sulphate, cuprous chloride, pyro--
gallol or phosphorous acid) may be separated from other elements
by distillation from a hydrochloric acid solution, the temperature of
the vapour being held below 10SPC; arsenic trichloride alone (also
germanium chloride, if ,present) volatilises and is collected in water
or in hydroch,loric acid.
B. Magnesium pyroarsenate, Mg~s201' Here the arsenic must

be present in the quinquevalent state (e.g., as arsenate). Phosphates
and other substances which give a precipitate with magnesia mixture
must be absent, as must also organic matter, for this reduces tne
substance and arsenic is volatilised.
C. Ammonium uranyl arsenate, NH.U0zAS04' xH20. and subse-
quent weighing as uranous uranate. U 308' The addition of a uranyl
salt solution to an arsenate solution containing excess of ammonium
ions results in the precipitatiQn of ammonium uranyl arsenate,
which is soluble in mineral acids but insoluble in acetic acid. Upon
igniting the precipitate, the arsenic is completely volatilised, leaving
a moss-green residue which consists mainly of U 30 S : this residue
is dissolved in concentrated nitric acid, and the resultant uranyl
nitrate upon cautious evaporation and subsequent ignition yields
pure black UaOs, and is weighed in this form (D. T. Lewis and V. E.
Davis, 1939).
If the solution contains arsenite, the latter must first be oxidised
to arseI1Jl.te with O·IN potassium bromctte in hydrochloric acid
solution at 70°C in the usual way (Section m.
137). A method is
thus available for the determination of arsenite and arsenate in ad-
mixture. The arsenite is first determined with standard potassium
bromate solution, and the total arsenate in the resulting liquid is
then determined by precipitation as the uranium salt.
A. Determination 01 arsenic as trisulphide. Procedure.
Pass a rapid stream of washed hydrogen sulphide through a
solution of the arsenious arsenic m
in 9N hydrochloric acid
at 15° to 20°C_ Allow to stand for an hour or two, and filter
through a weighed filter crucible (Gooch, sinteted glass or·
porous porcelain) (2). Wash the precipitate with 8N hydro-
chloric acid saturated with hydrogen sulphide, then succes-
sively with alcohol, carbon disulphide (to remove any free
sulphur which may be pr~sent) and alcohol. Dry at 105°C
to constant weight, and weigh as As 2S 3 •
Notes. 1. A suitable solution for eractice in this estimation is prepared by
dissolving about 0·3 g. of A.R. arsenious oxide, accurately weighed, in 9N
hydrochloric acid.
2. Sometimes a film of arsenious sulphide adheres to the glass vessel in
which precipitation was carried out; this can be dissolved in a little ammonia
snlution and the sulphide reprecipitated with the acid washing liquor.
B. Determination of arsenic as magnesium pyroarsenate.

Procedure. The solution must contain all the arsenic as
arsenate, and have a volume of not more than 100 ml. per
0·1 g. of arsenic. Add 5 ml. of conceptrated hydrochloric
acid slowly and with constant stirring to the arsenate.solution
(100 ml.) (1). Add a few drops of methyl red indicator,
followed by 10 to 15 ml. of magnesia mixture (2) for each 0·1 g,
qf As present; the magnesia mixture must be added drop-
wise and with continuous. stirring. Cool the solution in ice
water, and add pure concentrated ammonia solution slowly
and with constant stirring until the solution is alkaline. Stir
the liquid for a few minutes, and add 10 ml. of concentrated
ammonia solution. Allow the solution to stand at room
temperature for at least 4 hours, but preferably overnight.
Filter the precipitate through a weighed filter crucible (silica
Gooch or porous porcelain), wash with cold 1 : 25 ammonia
solution until free from chlqride, and then remove as much as
possible of the liquid by suction. Dry the precipitate at
100°C for 1 to 2 hours, place the crucible inside a nickel
crucible (Fig. 2-33), and apply heat slowly at first until no more
ammonia is evolved, and finally at the full temperature of a:
Meker type burner for 10 minutes (3). Cool in a desiccator
and weigh. Repeat the heating, etc., until constant weight
is attained. Weigh as Mg 2 As 2 0 7 (4).
Notes. 1. A suitable solution for practice is prepared by dissolving about
0·5 g., accurately weighed, of pure sodium arsenate Na 2 HAsO.,12H 20 (the
A.R. product is obtainable from the Mallinckrodt Chemical Works, U.S.A.;
see Appendix, Section A, 9) in 100 ml. of water. •
2. The magnesia minure is prepared as follows. Dissolve 25 g. of magne-
sium chloride MgCI 2,GH 20 and 50 g. of ammonium chloride in ~50 ml. of water
Add a slight excess of ammonia solution, allow to stand overnight, and filler
if a precipitate is present. Acidify with dilute hydrochloric acid, add 2 ml.
excess of concentrated hydrochloric acid, and dilute to 500 ml.
3. The best results are obtained by cOllductiag the final ignition in an
electrically-heated mume furnace at 800 0 to 900°C.
4. The following procedure (J. Dick, 1933) permits of the weighing as magne-
I\ium ammonium arsenate MgNH,AsO,,6H 20. Treat the concentrated arsen-
ate solution with 3 to 5 g. of ammonium chloride and 10 to 15 ml. of magnesia
mixture. Add hydrochloric acid dropwise until the precipitate which has
formed is dissolved. Add 1 drop of phenolphthalein to the cold solution,
stir, then enough 2·5 per cent ammonia solution to tum the solution red,
followea by an amount of concentrated ammonia solution equal to one-third
of the volume of the solution. At this stage the total volume of the solution
should not exceed 60 to 70 ml. for each 0·1 g. of As present. Cool in ice water.
for 1 to 2 hours, filter through a weighed crucible· (sintered glass or porous
porcelain), wash with 2·5 per cent ammonia solution, followed by 95 per cent'
alcohol (5 or 6 times), and small volumes anyhdrous ether (5 or 6 times).
Dry in a vacuum desiccator for 10 minutes or to constant weight. Weigh as
M&NH 4 AsO.,6H.O, which contains 25·89 per cent arsenic.
C. Determination of arsenic by precipitation as ammonium
uranyl arsenate and weighing as 'uranous uranate. Proced1tre.
The solution (150 ml.) should contain about 0·06 g. of As as
arsenate. Add 30 m!. of 4N ammonia solution, acidify with
acetic ,}cid, heat to boiling, and add 50 m!. (excess) of approxi-
mately O'lN uranyl acetate solution. Allow to stand several
hours, but preferably overnight; during this period the pale
yellow granular precipitate will become coarser. Filter
through a fine quantitative filter paper, wash free from soluble
salts, and transfer thefilter and precipitate to a weighed silica
or, better, a Main-Smith crucible. Heat in a fume chamber
provided with a good draught until all the carbon has burnt off
-the arsenic is simultaneOl!sly volatilised. Moisten the
residue with a few drops of concentrated nitric acid, and ignite
to constant weight over an ordinary Bunsen burner. Weigh
as UaOs.
IV,21. ANTIMONY
Discussion. Antimony may be determined in the following forms:
A. Antimony trisulphide, Sb~3' This method is of limited
application since no other elements that an! precipitable by hydrogen
sulphide in acid solution can be present, and the sulphide must be
dried and finally heated in an atmosphere of carbon dioxide at 280°
to 300°C. Arsenic can be separated by removal by distillation as
arsenic trichloride; tin can be removed by precipitation in the
presence of oxalic and tartaric acids or of phosphoric acid.
*Wash the crucible several times with alcohol (95 per cent), followed by
anhydrous ether, wipe the outside, leave in a vacuum desiccator for 5 minutes
and weigh.
518 .Quantitative Inorganic Analysis
B. Antimony tetroxide, Sb20 4 • Instead of weighing the antimony
as the trisulphidc after the procedure described in M elhod A, the
wet sulphide (or sulphide dried at 100°C) may be converted by
controlled oxidation with fuming nitric acid into the higher oxide,
and subsequently ~eatetl at 750 to 800°C, and weighed as Sbz0 4 •
0
Since tbis process is an inconvenient one and the oxide is very easily
reduced, it will not be described in detail.
C. Antimony pyrogallate, Sb(C sHs04)' Antimonious salts in the
presence of tartrate ions may be quantitatively precipitate.d with a
large excess of aqueous pyrogallol as the dense antimony pyro-
gallate. The method allows of a simple separation from arsenic ;
the latter element may be determined in the filtrate from the preci-
pitation of antimony by direct treatment with hydrogen sulphide.
The volumctric methods described in Chapter III are simpler,
quite as accurate and more rapid; they are therefore to be
preferred.
A. Determination of antimony as trisulphide. Procedure.
.
Quickly heat the solution of the antimony compound in
1: 4 hydrochloric acid (100 m!.) (1) contained in a conical.
flask to boiling and immediately pass a rapid stream of
washed hydrogen sulphide; maintain the solution a\ 90° to
100 e. Shake the flask gently at intervals afte]: the sulphide
0
has turned red, and keep the precipitate, as far as possible,

below the surface of the solution. As the precipitate darkens
in colour, reduce the gas stream. Continue the passage of
gas until the precipitate is crystalline- and black in colour
(total time required for precipitation is 30 to 35 minutes).
Dilute the solution with an equal v6lume of water, mix, and
heat again whilst lhe gas is slowly passed into the suspension
for some mi~utes. When the solution is clear, cool, and filter
through a filter crucible (Gooch, sintered glass or porous
porcelain) that has been heated at 280 0 to SOO°C and weighed.
Wash the precipita.te a few times with water to remove acid,
and then with alcohol; draw air through the crucible to dry
the precipitate lls far as possible. Place the crucible an.d
contents in a wide glass tube passing through the centre of an
air bath (Section n, 12F) or through an electrically-heated
tube furnace. Heat for 2 hours at 100° to l30°C in a current
of. carbon-dipxide (thi~ will completely dry the precipitate),
and then heat for a further 2 hours at 280° to 300°C (this pro-
cess will convert any Sb 2S5 present into Sb 2S 3 and will v.olatil-
ise the sulphur). Cool in a slow stream of carbon dioxide,
then place in a desiccator for 20 to 30 minutes;·and weigh as
Sb2 S a•
Note. 1. A solution, suitable for practice in this det;rmination, may bll
prepared oy diss_!)lving 0·5 g., accurately weighed, of A.R. anhydrous potas-
sium antimonyl tartrate or of pure tartar emetic in 150 ml. of 1 : 4 hydrochloric
acid.
C. Determination of antimony as antimonious pyrogallate.
Procedure. The solution should contain the antimony
(0·1 to 0·2 g.}.in the antimonious condition. Add a slight
excess over the calculated quantity of Rochelle salt (sodium
potassium tartrate) to avoid the formatiorr of basic salts upon
dilution. Dissolve approximately five times the theoretical
quantity of pure pyrogallol (Section 1,62,0) in 100 m!. of air-
free water, add this all at once to the antimony solution, and
dilute to 250 m!. After 30 to 60 seconds the clear mixture
becomes turbid, and then a dense cloudy precipitate forms
which separates out rapidly. Allow to stand for 2 hours,
filter through a weighed sintered glass or porous porcelain
crucible, wash several times with cold water to remove the
excess of pyrogallol (50 m!. is usually sufficient), dr}' at 100°
to 105°C to steady weight. Wash again with cold water,
dry at 100° to 10!)OC and weigh; repeat the operation until
the weight is constant. Weigh as Sb(C 6 H 5 0 4 ).
IV. 22. TIN •

Discussion. Tin may be determined in any of the follpwing ways:
A. Tin dioxide, 'Sn0 2 • The method depends upon the oxiqation
of the tin to the quadrivalent state, precipitation with ammonium
hydroxide, followed by ignition to. stannic oxide, Sn0 2 • This
procedure is satisfactory when applied to solutions containing tin
alone, but it cannot be used directly in the presence of many other
elements.
Tin is, h0>vever, frequently estimated as the dioxide in aHoys
(e.g., in brass or gun metal). The alloy is treated with' nitric acid,
when a precipit,ate of hydrous stannic oxide is formed; the formula
for this may be written, for simplicity, as H 2SnO a, mctastannic acid:
3Sn + H20 + 4HN0 3 = 3H2SnO a + 4NO.
The precipitate possesses colloidal character and exhibits power~ul
adsorption properties for certain ions, for example, iron, lead, cop-
per, nickel and zinc. It is advisable to use macerated filter pap'er in
the subsequent filtration and also to wash the precipitate with dilute
nitric acid in order to avoid peptisation. The precipitate is ignited
at the highest temperature of a M6ker type burner to stannic oxide:
H 2Sn0 3 = Sn02 + H 20.
Owing to the ready reduction of 'the oxide to the metal, the carbon
of the filter paper must be burnt off at as a Iowa t.emperature as
possible and the flame gases excluded from the crucible.
The precipitate"should be white, but rarely is because of the pres-
ence of impurities. The amount of impurity may be.determined by
adding to the impure weIghed oxide 15 times its weight of 'pure
ammonium iodide and heating for 15 minutes at 425° to 475°C, or
until no further fumes are evolved (E. R. Caley and M. G. Burford,
1936). The tin is volatilised quantitatively as stannic iodide:
SnO z + 4NH41 = SnI4 + 4NHa+ 2H zO.
The residue is treated with 2 to 3 m!. of concentrated nitric acid,
cautiously evaporated, ignited, and the residual metallic oxides
weighed; the loss in weight gives the weight of pure stannic oxide
present in the precipitate. The treatment with nitric acid is
necessary, for the other metals are converted into iodides and/or
oxyiodides, which are only slowly converted into the oxides by
heating in air.
B. Precipitati.on with cnplenon and weighing as stannic oxide,
SnOz• This process permits of the precipitation of tin in the pres-
ence of a.luminium, chromium, cobalt, nickel and manganese. In
the presence of 5 m!. of 48 per cent hydrofluoric acid per 200 to 400
ml. of solution, tin is not precipitated by hydrogen sulphide whereas
copper, lead, tervalent arsenic and antimony are thus precipitated;
many separations are therefore possible by this method.
C. Precipitation with phenylarsonic acid and weighing as stannic
oxide, SnOz• This method is applicable in the presence of zinc,
copper, and. other elements commonly associated with tin}n alloys,
etc. Zirconium and thorium seriously interfere and must be
abserlt.
The volumetric methods described in Chapter III are more· rapid
and equally accurate; they are therefore generally preferred.
A. Determination of tin as dioxide. Procedure. Weigh

out accurately 0·5 to 1·0 g. of the alloy (1) into a 150 m!.
beaker covered with a clock glass or Fisher" speedvap" beaker
cover. Add 15 mL of concentrated nitric acid mixed with 10
m!. of water while keeping the beaker covered. When the
vigorous action is over, evaporate the solution during a period
of at least 1 hour on a water bafh to a volume of 5 to 10 mI.,
but not to dryness. This digestion ensures the quantitative
precipitation of the stannic acid. Dilute to about 50 ml.,
and heat on the water bath fol' 30 minutes (to tlissolve soluble
salts). Add a Whatm·an· accelerator or one-quarter of a
Whatman ash less tablet, and stir with the hot suspension for
5 minutes. Filter through a Whatman No. 42 or No. 542
filter paper into a 150 m!. beaker; jf the first portion of the
filtrate is not clear, pass it through a second time, and add
more macerated filter paper if the precipitate persists in
Gravimetric Analysis 521'
running through. Wash the precipitate at least 10 times with
hot 1 per cent nitric acid. Fold up the filter, place it in a •
weighed porcelain crucible, dry and. char the filter as at a low
a temperature as possible, burn off the carbon at a very low
red heat to avoid reduction, and finally ignite the residue for
30 minutes at the full heat of a Meker type of burner whilst
keeping the crucible in a sfanding position, preferably in an
asbestos or " uralite " shield, to permit the free access of air.
(If reduction occurs, the precipitate may be moistened with a
drop or two Q.f nitric acid and ignited again.) Allow ~o cool
in a desiccator and weigh. Repeat the ignition, etc., until
constant weight is attained. Weigh as SnO z'
Ii the residue is appreciably coloured, the amount of
impurities may be determined as described in the Discussion.
Note. 1. Ridsdale's" Gun Metal, No. 6b" (one of the Analysed Samples for
Students) is suitable. Alternatively 0 ·15 to 0·2 g. of A.R. tin may be used.
B. Determination of tin with cupferron. Procedure. Remove
metals like copper, lead, tervalent arsenic and antimony, if
present, by precipitation with hydrogen sulphide in the pre-
sence of hydrofluoric acid (see Discussion), and boil off the
excess of hydrogen sulphIde. The solution should contain
0 . 1 to 0·3 g. of Sn and occupy a volume of 200 to 500 ml.
Add 5 to 10 tnl. of concentrated hydrochloric acid, 4 g. of
boric acid, 2 to.5 ml. of concentrated sulphuric acid, followed,
whilst stirring, with a liberal excess of filtered 10 per cent
aqueous cupferron solution. Stir vigorously for 30 to 45
minutes. The tin complex first separates in a white form
(? emulsion), rapidly assumes a curdy appearance, passes
through a stage when it resembles plastic sulphur, and finally
becomes compact and brittle, and may be crushed to a fine
powder with 'a g'lass rod. Filter upon a Whatman No. 41
or No. 541 filter paper, wash with cold water, dry in a weighed
crucible, expel the organic matter by gentle ignition, and then
ignite to constant weight. Weigh as SnO z.
C. Determination of tin with phenylarsonic acid,
C6HsAsO(OH)z. Procedure. Treat a weight of the sample
containing about 0·1 g. of Sn (1) with 10 ml. of I: I nitric
acid, heat gently to complete oxidation and until exce~s. of
nitric acid has been removed; take care not to dehydrate
the metastannic acid. Add concentrated hydrochloric acid
drop wise until complete solution is just obtained, dilute to
150 to 200 ml. with water, and adjust the acidity to 5 per cent
of hydrochloriC acid by volume (2). Heat the solution, and
s*
add 35 m!. of a saturated aqueous solution of phenylarsopic
_acid (:1). Allow to stand several hours, and filte~ through a
quantitative ftlte.r paper (4). Wash the precipitate with 4
per cent ammonium nitrate solution until free from chlorides
(or sulphates). Ignite the precip\tate and paper in a weighed
crucible over a small flame until all the carbon is burnt off,
and then heat in an electric furnace at 1075° to IlOO°C to
constant weight (5). Weigh as Sn0 2 •
Notes. 1. Vor practice in this determination, the student may use either
A.H. tin or Ridsdale's" Brass, No. 5b" or" Gunmetal, No. 6b."
2. If bismuth is absent, sulphuric acid (7'5 per cent by volume) may be
used.
3. The saturated aq~eous solution is approximately 10 per cent.
4. The addition of a little macerated filter paper (Section n, 29) aids the
filtration.
5. If an electric furnace is not available, a Meker or Fisher high temperature
burner miLy be used.
IV, 23. MOLYBDENUM

DisCll~sion. Molybdenum
, may be determined in the following
forms:
A. Lead·molybdate, PbMoO,. Precipitation as lead molybdate is
usually made by the slow addition of a..solution of lead acetate to a
hot dilute acetic acid-ammonium acetate solution containing the
molybclCnum. The method is applicable in the presence of copper,
mercury, cobalt, nickel, manganese, zinc and magnesium. Alkali
salts are not objectionable except sulphates, which must be absent
when the alkaline earths are present. A great excess.6f the preci-
pitant .should be avoided if chtorides are present. Free mineral
acids prevent ..complete precipitation, and elements such as iron,
chromium, .vanadium, tungsten, Silicon, tin arid titanium interfere'.
B. Precipitation with a,- benzoinoxime and weighing as molybdic
oxide, MoO a. Precipitation as the a-benzoinoxime complex is
made by the addition of excess of an alcoholic solution of the reagent
to a solution of the molybdenum compound containing 5 per cent
by volume of sulphuric acid. The precipitate cannot be weighed as
such, but must be ignited to molybdic oxide. The only oth"(']"
elemen'ts that give precipitates in mineral acid solution with the
reagent arc tungsten, palladium, sexiyalent chromium, quinque-
valent vanadium and tantalum. Chromium and vanadium do not
interfere if they are reduced to the tervalent and quadrivalent states
respectively with sulphurous acid. Tungsten may be determined
in the ignifed precipitate by extracting with the minimum volume
of warm 1 : 3 ammonia solution, filtering, acidifying the ammoniacal
solution with hydrochloric acid, and then adding a little cinchonine
solution; the solution is allowed to stand overnight, the precipitate
filtered off, washprl, ignited at 52GoC and weighed. The weight of
ungstic oxide is deducted from that of the molybdic oxide. -
C. Molybdenum "oxinate," IdoOz(CsHeON)z. Molybdates yield
sparingly orange-yellow molybdyl " oxinate " with oxine" solution;
the pH of the solution should be between the limits 3·3 to 7·6. The
complex differs from other" oxinates " in being insoluble in orgimic
solvents and in many concentrated inorganic acids. The freshly
precipitated compound dissolves only in concentrated sulphuric
acid and in hot solutions of caustic alkalis. This determination is
of particular interest as it allows a complete separation of molyb-
. denum and rhenium. _
, For t~~ volumetric determination of molybdenum, the precipitate
is brouglit into solution by rcpeated digestion with 2 to 4N sodium
hydroxide solution, the alkaline solution is slowly introduced with
constant stirring into 25 to 5{) m!. of concentrated hydrochloric acid;
the solution diluted to 50 to 100 mI., and then titrated with standard
O'lN potassium bromate (Section ill, 138; Procedure A).
I Ml. N KBr0 3 = 0·01199 g. Mo.
A. Determination of molybdenum as lead molybdate.
Procedur.;. Weigh out accurately about 0·4 g. of ferro-
molybdenp.m (1), dissolve it in 10 m!. of concentrated hydro-
chloric acid and 2 ml. of concentrated nitric acid, evaporate
to 2 to 3, mI., dilute to 50 mI., and transfer to a separating
funnel. Dissolve 5 g. of sodium hydroxide in 200 ml. of water
in the original beaker, heat to boiling, and wn in the solution
from the separating funner dropwise and with constant stir-
,ring. Rinse out the funnel twice with boiling water, and add
the washings to the main filtrate. Filter off the precipitatyd
ferric hydroxide and wash with hot water. Dissolve the pre-
cipitate ill the minimum volume of dilute hydrochloric acid,
and repreCipitate by slowly pouring into a solution of about
4 g. of sodium hydroxide in 100 ml. of water. Filter off the
precipitate, wash with hot water, and add the filtrate and
washings to the main filtrate. Acidify with acetic acid, add
50 ml. of a 50 per cent solution' of ammonium acetate, and
make up to 500 mI. in a measuring flask. Remove 250 m!. of
t.he solution (2), heat to boiling, and maintain near the boiling
point with a small flame; add from a burette a solution of
lead acetate {containing 4 g. of the salt and 1 ml. of glacial
acetic acid per 100 m!.) dropwise and with constant stirring.
When a slight excess of the precipitant has been added, the
milky solution clears appreciably. When this occurs, boil
for 2 to 3 minutes ,whilst the solution is stirred,.allow to settle,
and add a few drops of the reagent to see if precipitation is
complete. A large excess of precipij:ant should be avoided.
Digest on the steam bath for 15 to 30 minute~. Decant the
clear solution through a weighed Gooch or porous porcelain
crucible, wash by decantation.3 or 4 times with 75 m!. por-
tions of hot 2 to 3 per cent ammonium nitrate solution,
transfer the precipitate to the filter, and wash until the
soluble salts have b((en removed. Dry and ignite the pre-
cipitate at a dull red heat as described in Section IV, 140.
Weigh as PbMo0 4 •
Notes. 1. The Bureau of Analysed Samples" Ferro-Molybdenum, No. 200 ..
(a British Chemical Standard) is suitable. ,
2. If molybdenum is being determined in a simple salt, e.f!,., in A.R. molybdic
acid or molybdic anhydride, commence at this point. The solution should
contain about 0·1 g. of Mo in 200 m!. and may be prepared as follows. Dis-
solve 0·15 g., accurately weighed, of A.R. mQlybdic acid or anhydride, in (;0 ml.
of dilute ammonia solution, acidify with acetic acid, -add 25 m!. of a 50 per cent
solution of ammonium acetafe, and dilute to 200 n,ll.
B. Determination of molybdenum with a-ben~oinoxime
(cupron). Procedure. Prepare a solution containing 10 m!. of
concentrated sulphuric acid in a volume of 200 ml. and not
more than 0·15 g. of sexivalent molybdenum. If vanadates
or chromates are present, achl sufficient of freshly' prepared
sulphurous acid to reduce them, and heat to boiling: con-
tinue the boiling until the odour of sulp!J.ur dioxide can no
longer be detected. Copl to 5° to lODe, stir, and slowly add
10 ml. of a solution of the reagent (2 g. of a-benzoinoxime in
100 ml. of alcohol), and 5 ml. extra for each 0·01 g. of Mo pre-
sent (from 2 to 5 times the theoretical amount is required).
Continue the stirring, add just sufficient bromine water to
tInt the solution a ,pale yellow, and then add a few ml. of the
reagent. Allow the mixture to stand for 10 to 15 minutes at
10° to 15°C with· occasional stirring, and filter through a
Whatman No. 40 or 540 filter paper. Wash the precipitate
with 200 m!. of a cold, freshly-prepared wash solution (con-
taining 10 m!. of the reagent, 2 m!. of concentrated sulphuric
acid 'diluted to 200. mI.). Transfer the paper and washed
precipitate to a weighed platinum crucible, char the paper at
a low temperature without inflaming, and then ignite to
0
constant weight at 500 to 525 C. Weigh as Mo0 3 •
0
O. Determination of molybdenum with oxine. Procedure.

Neutralise the solution of· alkali molybdate, containing up to
0·1 g. of Mo, to methyl'red, and then acidify with a few drops
of 2N sulphuric acid. Add 5 ml. of 2N ammonium acetate,
dilute to 50 t() 100 mI., and heat to boiling. Precipitate the
molybdenum by the addition of a 3 per cent solution of oxine
in dilute acetic acid (for preparation, see Section IV, 3'iB;
Note 2) until the supernatant liquid becomes percepti-
bly yellow. Boil gently and'stir for 3 minutes, filter through
a filter crucible (sintered glass or porous porcelain), wash
with hot water until free from the reagent, and dry to
constant weight at 130° to 140°C. Weigh as MOO~(C9H60Nh.
IV, 24. SELENIUM AND TELLURIUM

Discussion. The gravimetric estimation depends upon the separa-
tion and weighing as elementary selenium or tellurium (or as tellu-
rium dioxide). Alkaliselenites and selenious acid are reduced in
hydrochloric acid solution with sulphur dioxide, hydroxylamine
hydrochloride, and hydrazine sulphate or ~ydrate. Alkali selenates
and selenic acid are not reduced by sulphur dioxide (they must first
be reduced to selenious acid by prolonged heating with concentrated
hydrochloric acid under reflux), but are readily reduced by hydrazine
sulphate or hydrate. In working with selenium it must be remem-
bered that appreciable amounts of the element may be lost on
warming strong hydrochloric acid solution_s of its compounds: no
appreciable loss occurs if dilute acid solutions are heated at tempera-
tures below 100°C.
With tellurium, precipitation of the element with sulphur dioxide
is slow in dilute hydrochloric acid solution and docs not take place
at all in the presence of excess of acid; moreover, the precipitated
element_is so finely-divided that it oxidises readily in the subsequent
washing process. Satisfactory results are obtained by the use of a
mixture of sulphur dioxide and hydrazine hydrochloride as the
reducing agent, and the method is applicable to both tellurites and
tellurates. Another method (B. S. Evans, 1933) utilises excess of
sodium hypophosphite in tha presence of dilute sulphuric acid as the
reducing agent. .
Two processes for the gravimetric estimation of mixtures of
selenium and tellurium are described. Selenium and tellurium
occur in practice either as the impure elements or as selenides or
tellurides. They may be brought into solution by mixing intim1 tcly
with 2 parts of sodium carbonate and 1 part of potassium nitrate in
a nickel crucible, covering with a layer of the mixture, and then
heating gradually to fusion. The cold melt is extracted with water,
and filtered. The elements are then determined in the filtrate.
1. Determination of selenium. ProcedJtre A. The selenium
must be present in the quadrivalent state and the seleniu~
content of the solution must not exceed 0·25 g. per 150 ~l.
Saturate the cool (15° to 20°C) selenium solution containing at
least 28 per cent by volume of hydrochloric acid (3 ·4N) with
sulphur dioxide, allow to stand until the red selenium sub-
sides, filter through a weighed filter crucible (Gooch, sintered
glass or porous porcelain), wash well successively with cold
concel1trated hydrochloric acid, cold water until free from
chloride, alcohol, and ether. Dry tJle precipitate for 3 to 4
hours at 30° to 40°C to remove ether, and then to constant
weight at 120 0 to 130°C. Weigh..as Se.
Note. A simpler method (J. Dick, 1930) of drying and weighing the precipi-
tated selenium, is to treat the well-washed precipItate, collected in a smtcred
glass Dr porous porcelain crucible, 5 or 6 times with small volumes of rectified
spirit and then similarly with anhydrous ether. The precipitate is sucked
drY' at the pump for 10 minutes, the crucible wiped with.Q clean linen cloth,
left in a vacuum desiccator for 10 minutes, and weighed.
Procedure B. The selenium may be present in either the
quadrivalent or sexivalent state. Dissolve the substance' con-
taining not more than. 0·3 g. of Se in 40 ml. of cold concen-
trated hydrochloric acid, and dilute to 100· ml. Add 10 ml.
of a 25 per cent aqueo~s solution of hydroxylamine hydro-
chloride, and heat at 90°C for 4 hours. Transfer the black
precipitate to a weighed filter crucible (Gooch, sintered glass
or porous porcelain), wash with water, then with alcohol, and
dry at llooe to constant weight.* Weigh as Se.
2. Determination of tellurium. Procedure A. The solution
should contain not more than 0·2 g. of Te in 50 ml. of 3N
hydrochloric acid (ca. 25 per cent by volume of hy.drochloric
acid). Heat to boiling, add 15 ml. of a freshly-prepared,
saturated solution of sulphur dioxide, then 10 m!. of a 15
per cent aqueous solution of hydrazine hydrochloride, and
finally 25 m!. more of the saturated solution of sulphur
dioxide. Boil until the precipitate settles in an easily filterable
form; this should require not m(){e. than 5 minutes. Allow
to _settle, fi.l.ter through a weighed filter crucible (Gooch,
sintered glass or ,porous porcelain). and immediately wash
with hot water until free from chloride. Finally wash with
alcohol (to remove all water and prevent oxidation). and dry
to constant weight at 105°C.* Weigh as Te.
In the alternative method of reduction, which is particu-
larly valuable for the determination of small amounts of
tellurium, the procedure is as follows. Treat the solution
containing, say, up to about 0·01 g. of Te in 90 mI..with, 10 ml.
of 1 :.3 sulphuric g'cip, then add 10 g. of sodium hypophos-
phite, and heat on a steam bath for 3 hours. Collect. and
weigh the precipitated tellurium as above.
Procedure B. Dissolve the material (elementary tellurium,
oxides or tellurites) in concentrated hydrochloric acid or in 10
per cent potassium hydroxide solution .. using 2 m!. per 0·2 g.
of oxide. If alkaline, just acidify the solution with hydro-
.. Alternatively use the rectified spirit-anhydrous ether method given in
Note to 1.
chloric acid, and dilute to 200 ml. with boiling water. Add a
slight excess of dilute ammonia solution, and just acidify
with acetic acid. Allow to cool, when the crystalline dioxide
will separate out. Collect the precipitate on a weighed filter
crucible (Gooch, sintered glass or porous porcelain), wash
with hot water, and dry at 105°C to constant weight. Weigh
as Te0 2 •
3. Determination of mixtures of selenium and tellurium.
Procedttre A. Dissolve the mixed oxides (n'ot exceeding 0·2fi g.
of each) in 100 m}, "of concentrated hydrochloric acid, and
add with constant stirring 50 ml. of cool concentrated hydro-
chloric acid which has been saturated with sulphur dioxide at
the ordinary temperatm;e. -Allow the solution to stand until
the red selenium has settled, filter through a weighed filter
crucible (Gooch, sintered glass or potous porcelain), and com-
plete the determination as described in 1, Procedure A.
Preserve the filtrate, hydrochloric acid and water washings.
Concentrate the latter on a water bath below 100 9C (above
100°C tel1urium is lost as chloride) to 50 mI., and estimate
the tellurium as described under 2, Procedure A.
Procedure B. Dissolve the mixed oxides (not exceeding
0'25 g. of each) in 40 ml. of concentrated hydrochloric acid.
Add 10 m!. of 25 per cent hydroxylamine'hydi-ochlo~ide, and
heat at 90°C for 4 hours. Filter off the black selenium through
a weighed filter cruCible, wash with water, then with alcohol,
and dry to cons~ant weight at 110°C. Determine the tellu-
rium in the filtrate exactly as detailed in 2, Procedlfre A.
IV, 25. PLATINUM

Discussion. Platinum may be determined in l'itllC'r of tlw following-
forms:
A. Metallic platinum. The platinum solution is treated with
formic acid, best at pH 6, and the precipitated platinum weighed.
B. Amnionium chloroplatinate, (NH4MPtC~], and subsequent
weighing as platinum. If platinum alone is present in hydrochloric
acid solution, it may be prccipitateu as ammonium chloroplatinatp
by the addition of a half-saturated solution of ammonium chloride,
washed, carefully ignited, and weighed at Pt.
A. Determination as metallic platinum. Pr(lcedllre. In this
estimation any excess of nitric and/or hydrochloric acid
present must be removed. Evaporate the solution of
platinum, containing no other platinum metals (ruthenium,
rhodium, palladium, osmium, and iridium) and gold, to a
syrup on the steam bath so as to remove as much hydro-
chloric acid as possible. If nitric acid was present, dissolve
the residue in 5 ml. of water, heat on the water bath for a 'few
minutes, ,add 5 ml. of concentrated hydrochloric acid, and
again evaporate to a syrupy consistency. Dissolve the residue
in water, and dilute so that the solbtion does not contain
more than 0·5 g. of Pt. in 100 ml. For each 100 ml. of solu-
tion, add 3 g. of anhydrous sodium acetate and 1 ml. of formic
acid. Heat on the boiling water bath for several hours. Fil-
ter through a quantitative filter papert Add a little more
sodium acetate and formic acid to the filtrate and digest in
order to ensure complete precipitation. Wash the precipitate
with water until free from chloride, dry. and ignite the filter
paper in contact with the precipitate (Section n, 36A) to
constant weight. Weigh as metallic Pt.
B. Determination of platinum by precipitation as ammonium
chloroplatinate. Procedure. Only hydrochloric acid may be
present in this estimation. The solution should .be free from
gold (which may b~ removed by the cautious addition of
sulphurous acid to the solution or of hydroquinone to the
solution 1·2N with respect to hydrochlori~ acid) and tlle
other platinum metals (~ee, however, below). If nitric acid
is presen1, it must be removed by several evaporations with
concentrated hydrochloric acid. Dissolve the residue in 5 m!.
of concentrated-hydrochloric acid and 20 m!. of hot water, and
then add gradually an equal bulk of P.alf-saturated ammonium
chloride solution. Allow to stand for some hours or over-
night. Filter off the precipitate, wash it with the ammonium
chloride solution, and finally twice with cold water. Transfer
the filter paper and precipitate to a Main-Smith crucible
(Section n, 36B), heat extremely slowly at first (otherwise loss
may occur), and ultimately raise to a strong red heat. Heat
to constant weight and weigh as Pt.
II other metals of the platinum group are present, the above
procedure does not give satisfactory results. It is necessary to'
proceed as follows. Rinse back the precipitate of ammonium
chloroplatinate into the precipitating vessel, and wash the filter
thoroughly with hot water. Slightly acidify the liquid (150 m!. for
I g. of Pt) with hydrochlpric acid, boil vigorously and keep stirred
by means of a glass tube through which a brisk current of chlorine
is passed (the operation must be performed in the fume cl1amber).
When the precipitate has dissolved (after a few minutes), stop the
current of chlorine, and wash the tube with hot water. The solution
now contains chloroplatinic acid. Then add just enough ammoniuIT!
chloride solution to combine with the platinum (6 m!. of a 10 per
cent solution for 1 g. of platinum), and evaporate (hot plate) to 20
m1. Allow to cool, and dilute with an equal bulk of saturated
ammonium chloride solution ... After a few hours, filter off the
coarsely crystalline precipitate, wash with half-sat.)lrated ammonium
chloride solution, then with a little cold water, ignite cautiously, and
weigh as Pt.
IV, 26. PALLADIUM

Diswssion. Palladium may be determined in ,one of the following
forms: ..
A. Palladotls cyanide, Pd(CN)2' and subsequent weighing as .the
metal. This is quite a good method particularly when large
quantities of .palladium are involved, because of the comparative
compactness of the.precipitate.
B. Palladium diIp.ethylglyoxime, Pd(C 4H,N 20 s)2' This is one of
the best methods for the estinr'd.tion of the element. Gold l11ust be
absent for it precipitates as the metal even from cold solutions.
The other platinum metals do not, in general, interfere. Moderate
amounts of platinum cause little contamination of the precipitate,
but with large amounts a second precipitation is desirable. The
precipitate is decomposed by digestion on the water bath with a
little aqua regia; and diluted with an equal vdlume of water; the
resulting solution is largely diluted with water, and the palladium
reprecipitated with dimethylglyoxime.
An objection to the precipitation of pal)adium ·with dimethyl-
glyoxime is the voluminous chaJ;:acter of the precipitate. Hence if
much palladium is present, an aliquot part of the solution should be
used, or the palladous cyanide method applied.
CO' Palladium a-nitroso-,B-naphtholate Pd(C,oH6 0 2N)2' a-NitrO'so-
,B-naphthol was first employed for the det_ermination of palladium
by precjpitating with this reagent and subsequently igniting the
slightly contaminated complex to, and weighing as, the metal. If
the hydrogen peroxide-sodium hydroxide method is employed
(compar:.e Cobalt, Section IV, 33A). the precipitate has the composi-
tion Pd(C 1oH s0 2N)2 and may be weighed in this form after drying
at 135°C. The method may be employed to separate palladium
from platinum.
D. Palladium a-nitro-,B-naphtholate Pd(C 10H,03N)2' a-Nitro-fiJ-
naphthol behaves similarly to a-nitroso-,B-naphthol, but is staBle in
aqueous solution. The precipitate is of definite composition,
Pd(CloH60~N2)2' and may be weighed after drying at 130 0 to 13Qo C.
The procedure provides a method of separation from platinum.
A. Determination of palladium as palladous cyanide. Pro-
cedure. This estimation must be carried out in a fume cup-
board provided with a good draught. Nearly neutralise the
530 Quantitative Inorganic
,
Analysis
solution containing up to 0·3 g. of Pd with sodium carbonate,
and add an excess of mercuric cyanide solution.' Heat on
the water bath 'until the odour of hydrocyanic acid has dis-
appeared, and allow the light yellow precipitate to settle.
Filter the precipitate through ~ quantitative filter paper.
Wash the precipitate first by decantation and then on the
filter with cold water. Dry and ignite in a weighed crucible
to constant weif;ht using the full heat of a Meker type of
burner. Weigh as Pd.
B. Determination of palladium with dimethylglyoxime.
Procedure. The solutidh should contain not more than 0·1 g.
of Fd in 250 ml., be 0'2QN with respect to hydrochloric or
nitric acid, and be free from nickel and gold. Add, at room
temperature, a 1 per. cent solution of dimethylglyoxime in
95 per cent alcohol. Use 2 to 5 ml. of the reagent for every
10 mg. of palladium. Allow the solution to stand for 1 hour,
and then filter through a weighed filter crucible (Gooch,
sintered glass or porous porcelain). Test the filtrate with a
little of the reagent to make sure that precipitation is com-
plete. Wash the orange-yellow precipita~e of palladium
dimethylglyoxime thoroughly, first with cold water and then
with hot water. Dry at 110°C to constant weight. Weigh as
Pd(C,H 7 0 2N 2) 2'
C. Determination of palladium. with a-nitroso-/3-naphthol.
Procedure. The solutIon should have a volume of about 20
ml. and contain about 0·03 g. of Pd. Treat the solution with
lO·drops of IOO-volume hydrogen peroxide, followed by 2N
sodium hydroxide until alkaline: cover the beaker with a
clock glass until the evolution of oxygen subsides. -Add 30
mL of glacial acetic acid, dilute to 150 mI., hea,t to boiling, and
precipitate with 15 m!. of the 1 per cent reagent solution (for
the preparation of the 2 per cent solution, see Section I, 62G).
Filter the reddish-brown flocculent precipitate through
a weighed sintered glass or porous porcelain crucible, wash 3
times with 20 per cent acetic acid, and finally with boiling
water until the washings are colourless. Dry to constant
weight at 135°C. Weigh as Pd(C 1o H 6 0 2N)2'
D. Determination of palladium with a-nitro-/3-naphthol.
Proct;dure. The palladium solution should have a volume of
about 20 ml. and contain not more than 0·03 g. of Pd. Add
10 drops of 100-volume hydrogen peroxide followed by 2N
sodium hydroxide until alkaline; cover the beaker with a
clock glass for oxygen is evolved. Dissolve the precipitat~
Gravimetrt'c A1~alysis 531
in glacial acetic acid and dilute so that the solu!ion contains
at most 5 per cent of acetic acid: a higher concentration of
acetic acid exerts a solvent action upon the precipitate. Pre-
cipitate with 15 ml. of the reagent (for preparation, see
Section I, 62H). Filter through a ,weighed sintered glass or
porous porcelain crucible, wash three times with 5 per cent
acetic acid, and finally with hot water. Dry to constant
weight at 130° to 135°C. Weigh as Pd(C10H~08Nk
IV,2.7. GOLD
Discussion. Gold is nearly always determined ,as the metal. The
reducing agents which were generally employed are sulphur dioxide,
oxalic acid and ferrous sulphate. No free nitric acid must be present
in the solution; if this is the case, the nitric acid must be removed by
repeated evaporation with concentrated hydrochloric acid, and the
solution diluted with water. With sulphurous acid, small amounts
of the platinum metals (particularly platinum) may be carried down
with the precipitate. It is therefore usually necessary to redissolve
the solid in dilute aqua regia and to reprecipitate the gold; oxalic
acid gives a better separation from the platinum metals in the second
precipitation, although the precipitate is somewhat finely divided.
Ferrous sulphate gives satisfactory results for gold·alone, but diffi-
culties are introduced if the platinum metals are subsequently to be
determined. •
The best results are obtained with hydroquinone as the reducing
agent (F. E. Beamish. J. J. Russell, and J. Seath, 1937). 'Precipi-
tation in hot 1·2N hydrochloric acid solutiotl is rapid, the gold is
readily filtered, and occlusion of the platinum metals is negligible.
Precipitation in the cold is complete in 2 hours. Palladium in the
filtrate can be precipitated directly with dimethyl-glyoxime, whilst
platinum in the filtrate may be determined by evaporating to dry-
ness in order to destroy organic matter and then digesting with a
little aqua regia.
Gold may also be separated from hydrochloric acid solutions of
the platinum metals by extraction with ether or with ethyl acetate
(compare Section 1,71); except in $pecia~ tases these methods do
not offer any special advantages over the reduction to the metal.
Determination of gold as the metal. Procedure A. The
solution should contain not more than 5 ml. of concentrated
hydrochloric aoid per -100 ml. of solution, not more than 0·5
to 1 g. of Au, and be free from lead, selenium, tellurium, and
the alkaline earths. Add 25 ml. of a freshly prepared satur-
ated sulPhur dioxide solption, and digest for 1 hour on the
steam bath in order to coagulate the precipitate. Add 5 to
.10 ml. more of the ?ul;phur dioxide solution, and allow to
532 Quantitative ! norganic Analysis
cool. ·If the cold solution smells strongly of sulphur dioxide,
the precipitation of gold is complete. Some of the metal is
finely divided, and it is therefore 'advisable to treat the filter
paper with a suspensipn of macerated filter paper prepared
from a Whatman accelerator or ashless tablet. Pour the
supernatant liquid through a Whatman No. 42 or No. 542
filter paper, preferably containing some filter paper pulp, and
transfer as little as possible of the precipitate to the paper
unless one precipitation is thought sufficient; this will only
be the case if very small amounts of platinum or palladium
are present. Wash well by decantation with hot dilute
hydrochloric acid (1 : 99). Transfer the filter to the beaker,
,and redissolve the gold in dilute aqua regia; use 8 m1. of
concentrated hydrochloric acid, 2 ml. of concentrated nitric
acid and 10 ml. of water for each gram or less of gold. Filter
from the paper pulp, and wash thoroughly with hot dilute
hydrochloric acid (1: 99). Evaporate the filtrate to dryness on
the water bath, add 2 to 3 m!. of concentrated hydrochloric acid,
and evaporate to dryness again ;. repeat this operation twice in
order to eliminate all the nitric acid. Treat the residue with 3 ml.
of concentrated hydrochloric acid, 5 drops of concentrated
sulphuric acid and 75 ml. of water, disregard the small amount
of gold which may separate, add 25 ml. of a saturated solution
of oxalic acid, and boil for a minute or two. If no further
visible precipitation of gold occurs, digest the solution on the
water bath for at least 4 hours. Filter off the gold through a
filter paper (as described above), and wipe the inside of the
beaker wuh" small pieces of quantitative filter paper to ensure
that all the metal is transferred from the beaker; wash well
with 1 : 99 hydrochloric acid. Transfer the filter to a weighed
porcelain or silica crucible, burn off the paper carefully, and
ignite to constant weight. Weigh as Au.
Procedure B. Treat the dilute solution of the gold, con-
taining only a little free hydrochloric acid, with a large excess
of a clear ferrous SUlphate solution, cover the beaker, and heat
on the water bath for several hours. Filter off the precipi-
tated gold through a filter paper, and complete the estimation
as under Procedure A.
Procedure C. The solution must be free from nitric acid, be
about 1·2N with respect to hydrochloric acid (ca. 5 m!. of
concentrated hydrochloric acid in 50 m!. of water), and con-
tain up to 0·2 g. of Au in 50 ml. H.eat the solution to boiling,
add excess of 5 per cent aqueous hydro quinone solutIon (3 ml.
for every 25 mg. of Au). and boil for ;W minut€s. Allow to
cool, and filter either through a weighed porous porcelain
crucible (A2) or through a Whatman No. 42 or 542 filter paper
(Section II, 28) ; wash thoroughly with hot water. The small
particles of gold remaining in the bottom of the beaker
(easily seen with a small flash lamp) are best removed with
pieces of ashless filter paper. Ignite the porous porcelain
,crucible to constant weight. If filter paper is qsed, transfer
to a weighed porcelain or silica crucible, and complete the
estimation as described in Procedure A.
IV, 28. ALUMINIUM

Discttssion. Aluminium ~ay be determined in the following
forms:
A. Aluminium oxide, A1 20 3 • Full details of this determination
(precipitation with ammonia solution) are given in Section IV, 9.
A more readily filterable precipitate may be obtained by precipita-
tion in weakly acid solution with a mixture of equal volumes of a
25 per cent aqueous solution of p~tassium iodide and a saturated
solution of potassium iodate. The mixture removes the acid pro-
duced hy hydrolysis, and causes the reaction to proceed from left to
right:
2AICl 3 + 3H20 ~ 2AI(OHh + 6HCI (i) ;
oHCI + KI0 3 + oK! = 3H20 +. 312 + 6KCl.
The If'cakly acid solution (200 to 300 m!. containing up to 0·2 g. of
AI) is treated with an excess of the K1-KI0 3 mixture; iodine
separates immediately. After 10 minutes the iodine is removed by
the addition of sodium thiosulphate solution with stirring from a
burette, and several m!. excess of the thiosulphate solution are
added. The mixture is heated on a water bath for 45 minutes with
frequent stirring. The hydrated awminium oxide is filtered off,
washed with hot water or with 2 per cent ammonium nitrate solu-
tion, ignited and weighed. The method may be applied also to the
precipitation of chromium and iron~ as hydroxides from dilute,
weakly acid solutions of their tervalent salts.
Sodium thiosulphate may also be employed to remove the acid
formed by hydrolysis of the' aluminium salt (equation (i) ) :
6HCI + 3Na 2S20 a = 3H2 0 + 6NaCI + 3S + 3S02 •
The mixture of aluminium hydroxide and sulphur, which is precipi-
tated, can be readily filtered and washed. Here the dilute solution
is neutralised with ammonium carbonate until a slight precipitate
, forms, which is dissolved in a little hydrochloric acid. Excess (say,
2 g.) of sodium thiosulphate dissolved in a small, volume of water is
added, the solution heated to boiling, and boiled until all trac;:es of
sulphur dioxide are expelled. 1: I Ammonia solution is added until
its odour is barely perceptible after blowing away the vapours, and
534 .Quantitative Inorganic Analysis
the mixture boiled for 2 minutes. TJ1e precipitate is flltered off,
washed, ,ilnd ignited in the usual way.
Precipitation of aluminium hydroxide can also be brought about
by bases other than ammonia, e.g., phenylhydrazine, pyridine, r
aniline, etc. Of these phenylhydrazine is sometimes employed for the

separation of alttminium from divalent iron in a form suitable for
immediate ignition and weighing as the oxide. (Phenyl hydrazine
also precipitates, best from solution of the sulphates, chromium,
titanium, zirconium, and thorium; beryllium is incompletely pre-
cipitated; ceric and ferric salts are partially reduced and incom-
pletely precipitated.) For the determination of aluminium the
dilute sulphuric acid solution, free from elements of Group II and
other interfering elements, is made slightly alkaline (methyl orange)
with ammonium h¥droxide solution, then just enough acid is added
to hold the other elements in solution, and the solution is diluted to
lOO to 200 ml. The whole is then heated, any iron pres<;nt is reduced
by the addition of a little saturated ammonium hisulphite solution,
the solution quickly neutralised with ammonia solution, and 6 to 7
drops of 1 : 1 hydrochloric acid added. One to three m!. of pure
phenylhydrazine is then added according to the amount of alumin-
ium to be precipitated, followed by one Whatman « accelerator" ;
the mixture is stirred until the precipitate becomes flaky. and is
then allowed to settle. The supernatant liquid ,vill be alkaline to
methyl orangE! but acid to litmus. If the precipitate has a brownish
colour, this is due to the colouring .matter contained in the phenyl-
hydrazine and not to ferric hydroxide. The precipitate is filtered
after a short time, washed with a solution of phenylhydrazine
sulphite,* and ignited to Al 20 s in the usual manner.
B. Basic aluminium succinate and subsequent 'igilition to the
oxide, Al20 a• Aluminium can be precipitated as the dense basic
~ccinate by boiling an acid solution containing succinic acid with
urea (H. H. Willard and N. K. Tang, 1937). The hydrolysis of the
urea forms ammonia gradually in a homogencpus solution, resulting
in apR of 4·2 to 4·6:
CO(NH 2}2 + RP' = CO 2 + 2NH a·
The dense precipitate is easily filtered and washed, and exhibits
much less tendency to adsorption of other salts t~an does the preci-
pitate obtained by precipitation as the hydroxide. Upon ignition,
the basic succinate is readily converted into alumina.
This method permits the separation of aluminium from large
amounts of ferrous iron, nickel, cobalt, manganese, zinc, calcium,
barium, and magnesium. For the separation of aluminium and
* Prepared by adding a cold saturated solution of sulphurous acid to a
few rol. of phenylhydrazine until the crystalline sulphite first formed dissolves
in the excess. and then adding phenylhydrazine drop-wise and with vigor~us
shaking until the odour of sulphur dioxide is no longer perceptible. Use 5 to
10 ml. of this solution in 100 ml. of hot water.
Gravimetric Analysis
..535
iron, the latter must be maintained in the ferrous state by the addi-
tion of phenylhydrazine; for zinc and aluminium, the precipitate
contains a little zinc which can be removed by volatilisation at 10000
to I 100°C in a current of hydrogen introduced through a silica Rose
crucible cover and inlet tube .
. C. Aluminium'S-hydroxy-quinolate, Al(C9 HoON)a. • Details of this
method are given in Section m. 138A. This procedure separates
.aluminium from beryllium, the alkaline earths, and magnesium.
For the gravime(ric determination a 2 or 5 per cent solution of oxine
in 2N acetic acid may be used; 1 m!. of the latter solution is suffi-
cient to precipitate 0·003 g. of AI. For practice in this determina-
tion, about 0·4 g. of A.R. ammonium alum, accurately weighed, may
be used'. This is dissolved in 100 ml. of water. The precipitated
aluminium "oxinate" is collected on a weighed filter crucible
(Gooch, sintercd glass or porous porcelain), well washed with cold
water, dried to constant weight at llO° to 120°C, and weighed as
AI(C 9 H oON)a'
B. Determination of aluminium as the basic succinate and

ignition to aluminium oxide. Procedure. The solution should
contain about 0·1 g. of Al and be acid with hydrochloric acid.
Add dilute ammonia solution until the solution becomes
slightly turbid, remove the turbidity with dilute hydrochloric·
acid, and add.! to 2 drops in excess. Add a sol.ution of 5 g.
of A.R. succinic acid in 100 m!. of water, followed by 10 g. of
ammonium chloride and 4 g. 6f urea ': dilute to CiOO m!. with
distilled water, Heat the solution to boiling and continue the
boiting for 2 hours after the solution has become turbid
(ca. 45 minutes).* Allow the precipitate to settle for a few
minutes, add a Whatman " accelerator," filter, and waSh 10
times with a 1 per cent succinic acid solution made neutral tb
methyl red with ammonia solution. If any precipitate ad-
heres to the sides of the beaker, dissolve it in a little dilute
hydrochloric acid, add a drop of methyl red or phenol red
indicator, and then dilute ammonia solution until just alka-
line; filter off the precipitate of aluminium hydroxide on a
separate small filter paper, and wash it with a.2 per cent
solutiun of ammonium nitrate. 'Place both papers and pre-
cipitates in a silica or, preferably, a platinum crucible, and
ignite to constant weight at 1200°C (compare Section IV, 9).
Weigh as A1 2 0 3 •
• The bOiling period (after the appearance of a turbidity) may be reuuced to
1 hour by first partially neutralising the hot solution to bromo-phenol blue or
to methyl orange by the drop-wise addition of dilute ammonia solution: <t
very faint opalescence will apfwar.
IV,29. BERYLLIUM
Discussion. Until about 1927 beryllium was nearly always deter-
mined by precipitation as the hydroxide Be(OH)2 with ammonia
solution in the presence of ammonium chloride or nitrate, and sub-
sequently weighing as the oxide BeO. The method was not entirely
satisfactory owing to the gelatinous nature of the precipitate, its
tendency to adhere to the sides of the v.essel, and its liability to
adsorption anq related effects. Beryllium hydfoxide was usualTy
separated with aluminium hydroxide, which it resembles in many
respects; two methods for the separation of beryllium and alumin-
iUl;n are given below. '
It has been shown {L. Moser, 1927) that.the hydroxide may be ob-
tained in an easily 'filterable and compact form by precipitatiC.l
from neutral solutions by heating with ammonium nitrate and methyl
alcohol at 70°C, a current of air being passed through the solution.
This method provides a separation from all metals of the later
analytical groups, exce,Pting barium, which should previously be
removed as sulphate; cadmium and thallium also do not interfere.
BeryHium may also be conveniently estimated by pr{!cipitation
with tannin in ammoniacal solution. Elements such as aluminium,
chromium, ferric iron, thorium, zirconium and titanium, are preci-
pitated by tannin from acetic acid solution; beryllium is not pre-
cipitated under these conditions, hence a good method of separation
•from these elements is available. If ferric iron is present, slight
reduction to the ferrous state occurs; this is prevented by the
addition of a few drops of IO-volume hydrogen peroxide solution
prior to precipitation with tannin.
Beryllium may also be quantitatively separated from aluminium
by means of oxine.
In all these methods the element is weighed as beryllia, which is
somewhat hygroscopic. The ignited residue must be cooled in a
desiccator over concentrated sulphuric acid or phosphorus pentoxide,
preferably in a tared and stoppered weighipg bottle, and weighed as
quickly as possible-usually after 25 minutes.
A. Determination of beryllium by precipitation with am-
monia solution. Procedure. The beryllium solution (200
mI.), pr~pared with nitric acid or hydrochloric acid and con-
taining about 0·1 g. of Be, must be almost neutral and contain
no other substance precipitable by ammonia solution. Heat
to boiling, and add dilute ammonia solution slowly and with
constant stirring until p,resent in very slight excess. Add a
Whatman accelerator or one-half of a Whatman ash less
• tablet, boil for I or 2 minutes, and filter on a Whatman No.
41 or No. 541 filter paper. Transfer as much of the precipitate
as possible by rinsing with hot 2 per cent ammonium nitrate
solution. Remove any precipitate adhering to the walls of
the beaker by dissolving in the minimum volume of hot very
dilute nitric acid, heating to boiling, and precipitating as
before. Filter through the same paper, and wash thoroughly
with the ammonium nitrate solution. Place the paper and
precipitate in a weighed silica, platinum, or Main-Smith
crucible, dry, heat at a low· temperature until the carbon
has been destroyed, and finally ignite with a Meker type of
burner (at approximately 1000°C). Allow to cool in a desic-
cator over concentrated sulphuric acid or phosphoric oxide,
and weigh rapidly after 25 minutes. Repeat the ignition
until constant weight .is attained. Weigh as BeO.
A'. Determination of beryllium by hydrolysis with ammo-
nium nitrite. Procedure. Neutralise the feebly acid solution,
containing about 0·1 g. of BeO per 100 ml., cautiously with
sodium carbonate, and remove the resultant ~light opan~s
cence with a few drops of dilute acid. Heat the solution to
70°C, pass a current of air through it, and add, with stirring,
50 ml. of 6 per cent ammonium nitrite solution and 20 mL of
methyl alcohol for each 0·1 g. of BeO expected to be present.
Any acid-forming oxides of nitrogen are thus removed as
methyl nitrite.' The solution becomes cloudy after a few
minutes. Complete the precipitation by boiling gently for
30 minutes. Add a further 10 mI. of methyl alcohol, and,
after 10 minutes, filter through a Whatman No. 41 or No. 541
filter paper, wash thoroughly with hot water, dry, and ignite
as in A. Weigh as BeG.
A". Determination of beryllium by means of tannin.
Procedure. The solution (300 to 400 ml.) should be weakly
acid, free from all cations other than those of alkali metals,
and contain at most 0·1 g. of BeO. Add 20 to 30 g. of am-
monium nitrate, heat to boiling, and treat with 10 per cent
aqueous tannin solution (use 10 times as much tannin as
there is BeO present) with con?tant stirring, followed by
ammonia solution, dropwise, until precipitation is complete.
(the solution should be just basic to litmus paper). The
precipitate is bulky but not gelatinous, is readily filtered,
and shrinks to a relatively small volume on suction. Filter
on a Whatman No. 40 or No. 540 filter paper resting upon a
Whatman filter cone (hardened, No. 51), and .apply suction.
Wash with hot water. If alkali metals are present, wash the
precipitate back into the beaker, dissolve in a little hydro-
chloric or sulphuric acid, and precipitate the boiling solution
with ammonia solution a,s before. FUter, dry the paper and
precipitate at llO° to 130°C, transfer to a weighed silica,
platinum or Main-Smith crucible, ignite, add a few drops of
concentrated nitric acid (to remove any organic matter which
may be present), ignite strongly, allow to cool in a desiccator,
and rapidly weigh as BeO. ,
·B. ,Determination of beryllium in the presence of AI. Cr.
Fe, Th and V by means of tannin. Procedure. The solution
should be neutral, occupy a volume of 300 to 400 ml. (depend-
ing upon the amount of aluminium, etc., present, but this
should not exceed 0·1 g.), and contain 30 to 40 g. of ammon-
ium acetate, and 20 to 25 g. of ammonium nitrate. Treat
with 15 to 20 ml. of 2N acetic acid, boil, and precipitate
by'the addition, with stirring, of excess of 10 per cent
aqueous tannin solution or of a 3 per cent solution of tannin
in saturated ammonium acetate solution. Filter the tannin
adsorption-complex on a filter paper with suction as in
A", and wash with dilute ammonium nitrate solutio'll.
Rinse back the precipitate into the beaker by means of
a. stream. of hot water from a wash bottle, and dissolve
it by the addition of dilute sulphuric acid to the hot liqu~d.
Nearly neutralise the resulting solution' with ammonia
solution, and repeat the precipitation as before with am-
monium acetate, ammonium nitrate and acetic acid. After
the precipitate has settled, Jccant the dear supernatant
liquid through the filter used 'for the first filtration. wash the
precipitate by decantation with hot ammonium nitrate
solJ.ltion, transfer it completely to the filter, and wash with
the hot ammonium nitrate soluti'on until the washings are
su!phate-free (test with barium chloride and hydrochloric
acid)'. Precipitate the beryllium in the filtrate and washings
by the addition of more tannin solution (if necessary), fol-
lowed by excess of ammonia solution. Complete the de~r
mination as in A". Weigh as BeO.
If ferric iron is present. add a lew drops of 10-volume hydrogen peroxide
before precipitating with tannin solution to minimise the chance of any ferrous
iron being formed.
If titanium and/or zirconium are present. precipitation may be effected in
the presence of more concentrated acid because the tannin complexes of these
metals are less soluble; only one precipitation is necessary. The procedure
is as follows. To the cold solution (300 mI.) add ammonia solution until it
produces an incipient cloudiness. then add ammonium acetate (10 g.). ammon-
ium nitrate (20 g.). and 80 per cent acetic acid (::!:; mI.). Heat the solution to
boiling. and precipitate the titania and lor Zirconia by the addition. with
stirring. of 10 per cent aqueous tannin solution in ten-fold excess of the weight
of the combined oxides. Boil for a few minutes. filter. wash the precipitate
with 10 per cent acetic acia and a little ammonium nitrate. Determine the
beryllipm in the filtrate and washings as above.
Note. If any onc of the above metals is present alone with the beryllium
it may be determined by igniting the tannin complex, produced in acetic acid
solution, lo th~ oxiue in the usual way.
B'. Determination 01 beryllium in the presence 01 aluminium

by means of oxine. Procedure. Precipitate the aluminium
> from an ammonium acetate ncetic acid solution (Section
IV, 28B), and keep the filtrate. Heat the filtrate to boiling,
and add a very slight excess of dilute ammonia solution. Filter
off the precipitated beryllium hydroxide (preferably after the
addition of macerated filter paper in the form of a Wh<rlman
accelerator or one-third of an ashless tablet), w4ich is coloured
yellow or b.t;own by adsorbed oxine, through a Whatman No.
41 or No. 541 filter paper, \vash thoroughly with a hot dilute
solution of ammonium acetate containing a few drops of
ammonia solution, and complete the determination as in A.
IV, 30. CHROMIUM

Discussion. Chromium may be determined in one of the following
forms:
A. Chromic oxide, Cr20 a• The chromium must be present in
solution as a chromic compound. It can be precipitated as the
hydroxide exactly as described under Aluminium in Section IV, 9
by means of the smallest possible excess of ammonia solution (the
precipitate is appreciably soluble in excesS of ammonia solution)
in the presence of excess of ammonium salts. It is important to
adjust the pH of the solution because of the amphoteric character of
hydrated chromic oxide; the colour of the solution prevents the
use of internal indicators so that electrcunetric methods must be
used; however, the indicator (methyl red) may be used" exter-
nally."
The KI-KIO a method (Section IV, 28) gives a more easily filterable
precipitate:
2CrCl a + 6H 20 ~ 2Cr(OH)a + 6HCl;
6HCI + 5KI + KIO a = 3H20 + 312 + 6KCl.
The best results are obtained by precipitation based upon the
slow hydrolysis of potassium cyanate (Darrington and A. M. Ward,
1930); this yields a granular precipitate of chromic hydroxide
which is readily filtered and w.ashed. This procedure will accord-
ingly be described in detail.
A disadvantage of the chromic oxide method is that the results
obtained are frequently a few tenths of a per cent too high on
account of the formation of small amounts of chromic chromate,
Cr2 (Cr04)3' for highly accurate results the precipitate should be
ignited and allowed to cool in hydroge~.
B. Mercurous chromate, Hg2Cr04 , and subsequent weighing as
Cr20 a. The chromium must be present as chromate, and the
540 Quantitatl've Inorganic Analysis
solution should be free from chloride and large amounts of sulphate.
Upon ignition, mercurous chromate is converted into chromic
oxide. Numerous kations interfere, hence the method has a very
limited application. -
C. Barium chromate, BaCr04 • The c4romium must be present as
chromate. The method is of limited application because of the
general insolubility of chromates. Chlorides do not interfere, but
sulphates must, of course, be absent. For further properties of
barium chromate, see Section IV, 45B. .
To tonvert a chromic salt into a chromate, treat the thromic
solution contained in a porcelain dish with several ml. of bromine
water, followed by freshly-prepared potassium hydroxide solution
until alkaline. Warm until the odour of bromine disappears.
To convert a chromate into 8: chromic salt, acidify the solution
with hydrochloric acid, and either add excess of freshly-prepared
saturated sulphurous acid and boil until all the sulphur dioxid~ has
been removed, or add concentrated hydrochloric acid and a little
alcohol, and boil until the odour of acetaldehyde can no longer be
detected and the solution is green. For practice in this operation,
determination A may be carried out with A.R. potassium dichromate,
previously dried at 180° to 200°C for 1 hour. 0·5 Gram of the salt
is accurately weighed out into a 600 ml. beaker, dissolved in about
150 mI. of water, 5 m!. of concentrated hydrochloric acid and 15 m!.
of alcohol added, and the whole heated until the solution is green
and no longer smells of alcohol or of acetaldehyde. The solution is
diluted to 300 mI., and the chromium precipitated as the hydroxide.
A. Determination of chromium as hydroxide by hydrolysis
of potassium cyanate.. Procedure. Weigh out accurately'
sufficient of the chromic salt to contain about 0·15 g. of Cr
(1), and dissolve it in 200 m!. of cold water: Add 5 g. 'of
ammonium chloride and 1 g. of potassiuHl cyanate, ~nd stir
the solution until the solid has dissolved. Heat the solution
slowly to boiling, when a fine granular precipitate separates
gradually. Allow the precipitate to settle, decant the turbid
liquor through a Whatman No. 41 or No. 541 filter paper (0),
and wash the precipitate once by decantation with hot 2·5
per .cent ammonium nitrate solution rendered faintly ammoni-
acal. Heat the filtrate to boiling, add a few drops of alizarin-S
indicator (2) followed by concentrated hydrochloric acid
until the solution is permanently yellow, and then add
ammonia solution until the indicator is just red. Collect the
small precipitate which is formed upon a separate filter (b), _
and wash it with hot ammonium nitrate solution. Wash the
main bulk of the precipitate by deoantation with ammonium
nitrate solution, filter off on paper (a), and wash .with the
GraiJimetric A nalysi§ 541
same solution. Dissolve the film of precipitate which
adheres firmly to the beaker in a few drops of hydrochloric
acid, reprecipitate it in the presence of alizarin-S by means of
·ammonia solution in very slight excess, and filter it upon the
second filter (b). Ignite both filters in a weighed platinum
crucible; when the carbon has burnt off, ignite for 10 minutes
at the maximum heat of a M6ker type of burner, and weigh
as Cr2 0 S '
Notes. 1. For practice, the student may employ about 2 g., accurately
weig!J.ed, of A.R. chrome alum KtSO"Cr2(SO.)a,24H20 for this estimation.
2. The alizarin-S indicator,is prepared by dissolving I g. of the solid in I
litre of water, and filtering from any insoluble matter present.
B. Determination of chromium by precipitation as mer-

curous cliromate. Procedure. The solution (1) should be
neutral 'or faintly acid, and contain about 0·1 g. of Cr as
chromate (or dichromate) in 100 m!. To the cold. solution
add a deCided excess of mercurous n.itrate solution (2),
whereby a brown amorphous precipitate of basic mercurous
chromate is formed. Heat the solution to boiling, stirring
frequently, until the precipitate is converted into the bright
red· and crystalline mercurous chromate, which s~ttles out
quickly; if the precipitation is complete, the supernatant
liquid will be colourless. Allow to cool, filter off the precipi-
tate through a What man No. 40 or No. 540 filter paper, and
wash it thoroughly with a cold 2·5 per cent aqueous solution of
mercurous nitrate. Dry the filter and precipitate at 100°C.
Transfer the filter and precipitate to a weighed porcelain or
platinum crucible, heat gently at first and afterwards with the
full Mcker flame in an oxidising atmosphere until constancy of
weight is attained. Weigh as Cr20 a• Owing to the poisonous
nature of the mercury fumes, the whole operation concerned with
the ignition of the precipitate must be carried out in a fume
cupboard provided with a good draught.
Notes. 1. A suitable solution for practice may be prepared by dissolving
0'5 to 0·6 g. of A.R. potassium dichromate, accurately weighed, in 250 m!. of
water.
2. Prepare the mercurous nitrate solution as follows. Moisten 3 :to 4 g.
of A.R. mercurious nitrate in a test tube with 2 drops of concentrated nitric
acid, add 10 ml. of water, and warm until the salt dissolves. Dilute the solu-
tion somewhat before use .•
C. Determination of chromium as barium chromate. Pro-

cedure. The solution should contain about 0·1 g. of Cr as
chromate, be. neutral or weakly acid with acetic acid, and
occupy a volume of 200 to 3QO m!. Add a 10 per cent solution
of barium acetate dropwi.se from a burette and with constant
542 Quantitqtive Inorganic Analysis
stirring to the boiling solutioR (1) until present in slight excess.
Place the beaker on a water bath until the precipitate seUles ;
test for completeness of precipitation by adding a little more
of the reag~nt. Allow to cool, filter through a weighed Gooch
or porous porcelain crucible, wash with hot water until 1 ml.
of the washings gives no precipitate with a little dilute sul-
phuric acid. Complete the determination as described in
Section IV, 45B. Weigh as BaCrO,.
Note. 1. A solution for practice may be prepared by dissolving about 0'5 g.
of A.R. potassium dichromate, accurately weighed, in 300 m!. of water, addi,p.g
ammonia solution until neutral, and then 1 m!. of.6N acetic acid.
IV. 31. Iron.

Discussion. Iron may be determined in the following forms:
A. Ferric oxide, Fe 20 a• The precipitation of iron as ferric-hydrox-
ide by ammonia solution, etc., and its conversion into ferric oxide,
in which form it is weighed, is fully described in Section IV, 8.
Precipitation with ammonia solution yields a gelatinous precipitate
which is somewhat difficult to wash and to filter ;. this difficulty is
partly overcome by the use of macerated filter paper.
There are, however, three methods of precipitation which yield
ferric hydroxide in a relatively dense and granular form, which is
easily washed and filtered. These are:
(i) The potassium cyanate method. This has already been dis-
cussed under Chromium, Section IV, 30.
(ii) The hexamethylenetetramine method. This is based upon the
slow hydrolysis of hexamethylenetetramine in acid solution:
2FeCI a + 6H 20 ? 2Fe(OH}a + 6HCI; _,.
(CH2)6N4 + lOH 29 ~ 6HCHO (formaldehyde) + 4NH 40H.
Separation of ferric iron is effected from manganese, zinc, cobalt and
nickel, but not from aluminium, chromium and titanium, which are
precipitated. If calcium is present, there is no risk of coprecipita-
tion as carbonate, as when ammonia solution is used, for the latter
reagent frequently contains traces of carbonate.
(iii) The hydrazine hydrate method. This procedure is particu-
larly valuable in the separation of ferric iron from manganese.
Precipitation must be carried out .in neutral or slightly alkaline
solution, for in acid solution reduction to the ferr.ous state occurs.
Full experimental details for each of these methods will be given.
Attention is directed to the fact that if ignition is carried out in a
platinum crucible at a temperature above IlOO°C some reduction
to triferric tetroxide Fe304 may occur, and at temperatures above
1200°C some of the oxide may be reduced to the metal and alloy
with the platinum. This accounts in part for the contamination of
the platinum crucible by iron which sOIl}etimes occurs in analytical
GraV1:metric Analysis 543
work. No magnetic o~ide of iron is produced if Main-Smith
crucibles (Section n, 36B) are employed for the ignitions.
B. Ferric cupferron and subsequent weighing as ferric oxide.
Cupferron, the ammonium salt of nitrosophenylhydroxylamine,
CsH sN(NO).0NH 4 , precipitates iron, vanadium, titanium and
zirconium from 1ttrongly acid sol}ltions, thl\S affording a separation
from aluminium, chromium,. beryllium, phosphorus, boron, man-
ganese, zinc, nickel, cobalt, and sexivalent uranium. Copper and
thorium must be precipitated from weakly acid solutions. Several
metals (e.g., lead, silver, mercury, bismuth, tungsten and cerium)
interfere, but most of these may usually be removed by other
methods-for example, by hydrogen sulphide in acid solution.
The precipitate cannot be weighed as such, but must be ignited to
the oxide. The ignition must be very carefully carried out in the
early stages in order to avoid mechanical losses, for wet precipitates
tend to liquefy and effervesce, whilst <)ry precipitates give off
considerable volatile matter. The precipitate is rather bulky, and
the amount of material taken should therefore be such as to yield
O·I.to 0·2 g. of ferric oxide.
A. Determination of iron as hydroxide by hydrolysis of
potassium cyanate. Procedure. The solution (200 ml.)
should contain 0 ·15 to O' 20 g. of Fe in the ferric state. Add a
few drops of concentrated hydrochloric acid, and heat to
boiling (1). Add a concentrated aqueous solution containing
2 g. of potassium cyanate to the boiling solution as rapidly as
possible, with vig9rous stirring (2). Boil the solution for 2 to
3 minutes, and allow the precipitate to settle. Wash well by
decantation with hot water, filter off, and complete the deter-
mination as'in Section lV, 30A. It is better to ignite the wet
filter paper and precipitate in a Main-Smith crucible. Weigh
as Fe2 0 s'
Notes. 1. If the yellowish solution darkens during heating, add more
hydrochloric acid to bring back the solution to the original colour.
"2. The cyanate solution cannot be added all at once, for frothing will then
be excessive. It is important that the ferric hydroxide sol stage be passed
through as quickly as possible.
A'. Determination of iron as hydroxide by means of
hexamethylenetetramine. Procedure. The solution (200
mI.) should be faintly acid and contain about 0·1 g. of Fe
in the ferric state. To the cold solution add 20 g. of ammo-
nium chloride, followed by excess pf a 10 per cent aqueous
solution of hexamethylenetetramine with stirring. Allow the
precipitate to settle, decant the mother liquor through a
Whatman No. 41 or No. 541 filter paper, wash the precipitate
by decantation with hot water, transfer-to the filter (I), and~
again wash with hot water. Ignite the precipitate and
weigh as, FezO a. '
Note. 1. If a tenacious film of the precipitate adheres to the beaker, dissolve
it in a few drops of concentrated hydrochloric acid, and reprecipitate with
ammonia solution.
A". Determi.n:ation' of iron as hydroxide by means of

hydrazine hydrate. Procedure. The solution (100 ml.)
should contain ~bout 0·1 g. of Fe in the ferric state. Add 8
g. of ammonium chloride, a few drops of methyl red indicator,
and neutralise with ammonia solution (methyl red turns
yellow). Then add in one operation a slight excess of hydra-
zine hydrate solution (I), and boil for a short time. Filt~r
through a Whatman No. 41 or No. 541 filter paper, wash with
hot water containing 1 g. of ammonium chloride and 1 ml. of
hydrazine hydrate per 100 ml., and then with hot water.
Dry and ignite in the usual way. Weigh as Fe 2 0 3 •
Note. 1. Prepared by adding 4 m!. of hydrazine hydrate (poisonous) to
100 m!. of water.
B. Determination of iron with cupferron. Procedure.

The solution (1) (150 to 200 ml.) should contain about 0·1 g.
of Fe in the ferric state and be strongly acid with hydro-
chloric acid or sulphuric acid. To. the cooled solution (ca.
10°C) add a freshly-prepared-, filtered 5 per cent aqueous
cupferron solution (2) slowly and with constant and vigorous
stirring until no further formation of a brown precipitate
takes place. The formation of a white precipitate of nitroso-
phenylhydroxylamine indicates when the reagent is present
in excess. Do not warm -since the reagent is rapidly decom-
posed in hot acid solution. Add a Whatman accelerator (or
a third of a Whatman ashless tablet), stir for 2 to 3 minutes,
and without further delay filter through a Whatman No . .41
or No. 541 filter paper. Wash several times with 10 per cent
by volume of hydrochlq,ric acid containing 1·5 g. of cupferron
per litre, then twice with 5N ammonia solution to remove
excess of cupferron, and finally once.with water. Ignite the
precipitate w~th the paper in a weighed porcelain, silica or
platinum crucible (best in a Main-Smith crucible, Section
n,36B) very gently at first until all 'the organic matter is des-
troyed, and then strongly to constant weight. Weigh as
Fe20 s·
Notes. 1. A suitable solution fpr practice may be prepared by weighing out
accuratelyO·g to 0·9 g. of A.R, iron alum, dissolving it in 150 m!. of water, and
adding 25 m!. of concentrated hydrochloric acid.
Gravimetrt'c Analysis Mi>'
2. Only freshly-prepared solutions should be employed, because the solution
only keeps for a day or two. The dry reagent should be kept in a cool dark
place, and preferably with a bag of ammonium carbonate suspended in the
bottle.
IV, 32. NICKEL

Discussion. Nickel may be determined in one of the following
-forms:
A. Nickel dimethylgiyoxime, Ni(C4H702N2)~' Full details .of this
estimation are given in Se~tion IV, 12.
B. Nickel salicylaldoXime, Ni(C 7H60 2N)2' The complex is precipi-
tated in neutral or very faintly acid s91utions (best at PH = 7) in
contrast to that of copper (Section IV, 19C), which is formed in the
presence of acetic acid. The experimental details are similar to
those already given for Copper, except that the solution must be
neutral or very faintly acid. Experimental details for the stWaration
of copper and nickel with salicylaldoxime are given below.
C. (Ni(CsHsN)J(CNS)2; pyridine method. This is' a rapid method
similar in advantages and limitations to those described for copper
-(Section IV. 19E). cobalt (Se<:tion IV. 33D) and zinc (Section
IV,3ID). A determination may be completed in about 30 minutes.
D. Metallic nickel. Experimental details of the electrolytic method
are given in Section IV,S1.
B. Determination of nickel in the presence of copper with
salicylaldoxime. Procedure. Treat the solution, free from
mineral acid and containing both nickel (not more than 0·03
g.) and copper (about O;.03_g.), with 1 g. of sodium acetate and
10 mI. of glacial acetic acid per 100 mI. of solution. * Add
excess of salicylaldoxime reagent over the quantity required
to precipitate both metals, and stir the solution vigorously
during the addition. Filter off the precipitated copper
~omplex on a weighed fil~er crucible (Gooch, sintered glass, or
porou~ porcelain), wash w.ell with cold watert and dry at 100°
to lO5°C to constant weight. Weigh as C,u(C,H S 0 2 Nh.
Calculate the weight of copper present. Add dilute ammonia
solution to the filtrate and washings (diluted to 400 to 350
mls.) until th& solution is very faintly acid or alkaline.
Stir thoroughly to coagulate the precipitate of nickel salicylal-
doxime. Filter through a weighed sintered glass or porous
porcelain crucible, wash with cold water until the washings
give no·colouration with ferric chloride solution, and dry at
100°C to constant weight. Weigh as Ni(C7 H 6 0 2N)2- Calcu-
late the weight of nickel present.
• The best pH range for the precipitation of copper in the' presence of.nickel
is 2·6 to 3·1.
T
• 546 Quantitative Inorganic Analysis
C. Determination of nickel by the pyridine· method. Pro-
cedl/re. The solution (100 m!.) should contain 'about 0·1 g.
of nickel (1) and be n~utral in reaction. Stir in.0·5 to 1·0 g.
of A.R. ammonium thiocyanate, heat to boiling, add 1 to 2
ml. of pure pyridine, and immeqiately remove the flame.
Stir for 2 to 5 seconds until the precipitate commences to
separate in sky-blue prisms. (The precipitate separates
immediat,ely or after standing a short time according to the
quantity of nickel present.) When cold, filter through, a
weighed sintered glass or porous porcelain crucible, and
use solution 1 to assist in the transfer of the precipitate
to the crucible. Wash 4 to 5 times with solution 2, then
twice with 1 m!. portions of solution 3, and finally 5 or 6
times with 1 m!. portions solution 4. Dry in a vacuum
desiccator at room temperature for 10 minutes and weigh.
Repeat the drying until the weight is constant. Weigh as
[Ni(CsHsN)J{CNS)2 (2).
Solution 1. 100 Ml. of water containing 0·4 g. of NH 4CNS and
0·6 ml. of pyridine.
Solldion 2. 61·F "vn. of water, 37·0 ml. of 95 per cent alcohol, 0·1 g.
of NH 4 CNS, and 1·5 ml. of pyridine.
Solution 3. 10 Ml. of absolute alcohol and 0·5 rol. of pyridine.
Solution 4. 20 Ml. of absolute (sodium dried) ether and 2 drops of
pyridine.
Notes. 1. For practice in this estimation, the student may employ 0·3 g.
accurately weighed, of A.r... nickel sulphate. " -
2. For further experimental details, see under Copper and Zinc.
IV, 33. COBALT

Discussion. Cobalt· may be separated in olle of the following
forms:
A. Cobalti a-nitroso-,B-naphtholate, Co(ClOHs02N) 3,2H20.
a-Nitroso-p-mtphtholgives a red precipitate with solutions of cobalt
salts. The composition is substantially that of the cobalti-com-
pound, but small quantities of impurities (probabJy containing some
of the cobaltous compound). are present, and this renders the results
somewhat inaccurate. However, by converting the cobaltous salt
into cobaltic hydroxide with hydrogen peroxide and sodium
hydroxide, and then precipitating the solution of cobaltic IrYdroxide
with the reagent, the pure cobalti-a.-nitroso-p-naphtholate Clihydrate
is obtained, and may be weighed in this form after drying at 130°C.
The reagent is unstable in aqueous solution, and must be freshly
prepared as required. For the limitations of the method, see
Section I, 62F. .
Gravimetric A,.nalysis 547
~. Cobalti a-nitro-,B-naphtholate, Co(CloHsOaN)a. a-Nitro:,B-
naphthol reacts similarly to a-nitroso-,B-naphthol, but has the ad-
vantage that its aqueous solution is stable. The precipitate ob-
tained by the hydrogen peroxide-sodium hydroxide method is
homogeneous, Co(C 1oH6 0 3N)a, and may be weighed as the cobalti
complex after drying at 130°C. •
C. Cobalt phenylthiohydantoate complex. Phenylthiohydantoic
acid precipitates cobalt in ammoniacal solution in the presence of
citrate as a red-brown complex of somewhat indefinite composition.
The precipitate may be ignited to the oxide" Co a0 4 " of uncertain
composition, the latter fused with potassium pyrosulphate, and the
cobalt in the aqueous extract determined by any of the other methods
given in this Section. For the limitations and advantages of the
method, see Section I, 62H. .
D. [Co(CSH 5N)4] (CNS) 2; pyridine method. If a dilute neutral
solution of a cobalt salt is treated with ammonium thiocyanate and
pyridine, tetra-pyridine cobalt dithiocyanate is quantitatl'vely pre-
cipitated. The precipitate is collected and washed inter alia with
alcohol and ether containing a little pyridine, and then dried by
simply leaving in a vaCuum desiccator for 5 to 15 minutes. A
complete determination of cobalt may. be carried out in about 30
minutes. The precipitate may be kept unchanged ior several
hours in a vacuum desiccator. The method can be applied in the
presence of the alkali metals, magnesium, and the alkaline earth
metals, but copper, cadmium, nickel, manganese and zinc must be
absent. Large quantities of ammonium salts must be absent since
these exert a slight' solvent action upon the pr.ecipitate. If the
solution is strongly acid, it should be evaporatecl to dryness and the
residue dissolved in water.
E. Metallic cobalt. The electrolytic method is discussed in
Section IV, 82.
A. Determination of cobalt with a-nitroso-,B-naphthol.
Procedure. Concentrate the weakly acid solution of the
cdbaltous salt, containing not more than 0·03 g. of Co(l), to
10 to 20 ml., and allow-to cool. Treat the cold solution with
5 to 10 drops of 100-volume hydrogen peroxide, and tilen add
2N sodium hydroxide to precipitate the cobalt as cobaltic
hydroxide Co(OH)a; keep the beaker covered duriflg this
proces~ since oxygen is evolved. Add .10 to 20 ml. of glacial
acetic acid, warm gen~ly to dissolve the precipitate, dilute to
200 ml. with boiling water, and treat with 10 to 20 ml. of the
freshly-prepared reagent (for preparation, see Section I, 62F) ;
about 0·25 g. of a-nitroso-,B-naphthol is required for every 0·01
g. of Co. . Heat to gentle boiling with vigorous stirring unt_il
the· precipitate coagulates or settles out. The supernatant
liquid should be clear and yellow. Test whether precipitation
is complete by adding a little more of the reagent to the clear
solution. Filter through a weighed Gooch, sintered glass or
porous porcelain crucible, transfer the precipitate to the
crucible with the' help of a little 33 per cent acetic acid, and
wash with this hot acetic aQid. Finally wash the. precipitate
with boiling water, and dry at 130°C to constant weight.
Weigh as Co(C]oH 6 0 2N)3,2H 20.
Note. 1. A suitable solution for practice may be prepared from ca. 0·1 g.,
accurately weighed, of A.R. cobaltous sulphate or A.R. cobalt ammonium
sulphate.
B. Determination of cobalt with a-nitro-,B-naphthol. Pro-
cedure. The method is similar' to that described under A.
Concentrate the cobalt solution, wealHy acid with mineral
acid and containing not more than 0·03 g. of Co, to 10 to 20
m!. Treat the cold solution with 10 drops of lOO-volume
hydrogen peroxide, then add 2N sodium hydroxide to preci-
pitate the cobaltic hydroxide, and cover with a clock glass.
Dissolve the precipitate by the addition of 10 ml. of glacial
acetic acid, warm gently if necessary, and dilute to 150 to 200
ml. with hot water. . Add a 50 per cent excess of the reagent
(for preparation, see Section I, 62G) slowly and with constant
stirring. Heat, with stirring, to the boiling point, allow to
settle, and filter through a weighed sintered glass or porous
porcelain crucible. Use a little 30 per cent acetic acid to
assist in the transfer of the precipitate. Wash 3 times with
30 per cent aceti<;; acid, followed by several tim:es with hot
water, and dry at 130°C to constant weight. Weigh as
Co(CloH60aN) 3'
C. Determination. of cobalt with phenylthiohydantoic acid.
Procedure. The cobaltous solution (400 ml.) should contain
not more thaI! 0 ·025 g. of Co anQ. about 7 g. of citric acid.
Add concentrated ammonia solution until the solution fis
neutral and then a further 10 ml. -Add 0·7 to 1·0 g_ of
phenylthiohydantoic acid, dissolved in 30 ml. of hot alcohol,
with vigorous stirring. Boil and stir the solutio!! for a few
minutes to cause the precipitate to rise to the surface of the
liquid. Filter the bulky precipitate through a Whatman No.
41 or No. 541 filter paper, wash well- with a 0·5 per cent
ammonium citrate solution, and dry at 100° to 105°C. The
mater.ial thus obtained is not pure cobaltous phenylthiohy-
dantoate, and upon ignition it gives impure CoaO,; neverthe-
less, if';t is ignited in a porcelain crucible over a Meker type
burner, fairly accurate results are obtained by calculating as
if it were C0 20 3 • However, it is usua,lly best to fuse the impure
Gravimetric Analysis f)49
'oxide with pure potassium pyrosulphate, disselve the cooled

melt in water, and determine the cobalt by methods A, B, D
or E.
D. Determination of cobalt by the pyridine method. Pro-
t;ed II re. The solution (75 to 100 ml.) should be neutral and
~ontain about 0·1 g. of CO (I). Add 0·5 to 1·0 g. of A.I<.
ammonium thiocyanate, heat to boiling, treat with 1 to 2 ,
ml. of pure pyridine, and remove the "Source of heat imme-
diately. Stir for 5 s~conds, and allow to cool to room tem-
perature. The solution remains clear for a short tim~; as
the solution cools, red shiny crystals of the complex separate.
(If the precipitate on the sides of the beaker exhibits the blue
colour of the di-pyridine complex, this is washed with a little
water into the solution when the blue colour immediately
disappears. If stirring is continued for se>;eral minutes after
the addition of the pyridine, the compound soon separates in
tl1e crystalline form.) Filter through a weighed sintered glass.
or porous porcelain crudble; use ~ollttion I to assist in tM
transfer of the precipitate to the crucible. Wash the preci-
pitate .5 or 6 times with solution 2, then twice with 1 m!.
portions oj solution 3, and finally 8 to 10 times with small
volumes (1 to 2 ml.) of solution 4 (for further experimental
details~ see Sections IV, 19E and IV, MD). Dry in a vacuum
desiccator at [Dom temperature for 10 minutes, and weigh.
Repeat the drying, 1'!tc., until the weight is constant. Weigh as
[Co(CsHsNM(CNSh·
Solution 1. 100 MI. of water containing 0·5 g. of NH 4CNS and
0'7 m!. of pyridine.
Solution 2. 85·5 MI. of water, L3'O ml. of 95 per cent alcohol.
O·l.g. of NH 4CNS -and 1'5 m!. of pyridine. .
Solution 8. 25 Ml. of absolute alcohol and 2 ml. of pyridine.
Solution 4. 30 1\11. of absolute (sodium dried) ether and 4 drops of
pyridine.
Note. 1. For practice in this estimation, the student may employ 0'3 g.,
accurately weighed or of A.R. cooalt,.al}lI11onium·suJphate of A.R cobaltous
sulPhate. ,
If the solution is weakly acid, add ammonium thiocyanate, then pyridine
until a precipitate appears, warm until the precipitate dissolves, add 1 to 1·5
mL of pyridine, and immediately remove the source of heat.
IV, 34. ZINC

Disczession. Zinc may be determined in any of the following forms;
A. Zinc ammonium phosphate, ZnNH4P0 4, or zinc pyrophosphate,
Zn2PZ0 7 • The zinc is precipitated, with close adherence to the
experimental conditions detailed below, as zinc ammonium. phos-'
phate ZnNH,P0 4 ,6H 20 and weighed either as ZnNH 4P0 4 (after
drying at 105°C) or as Zn 2P 20 7 (after ignition at 900° to 1000°C).
The method is of limited application because of the numerous other
elements which form sparingly soluble phosphates.
The precipitant is diammonium hydrogen phosphate (NH.)2HPO.,
which exerts an important buffering action:
ZnCl z + (NH.hHPO. = ZnNH,P0 + NH.CI + HCI;
4
HCI + (NH4)2HPO. = NH 4H 2 PO. + NH4Cl.
It has been shown that the most satisfactory results are obtained
over the pH range 6·4 to 6·9 (optimum PH is 6·6), and that when the
pH exceeds 7·0 the precipitate is removed from the beaker with
difficulty. A solution containing 5 to 10 g. of ammonium chloride,
10 ml. of 2N sodium acetate and 10 ml. of 10 per cent (NH.)2HP04
solution in 150 ml. has, a pH of 7·5; this automatically falls to 6·4
to 6·9 when heated for 2 hours on the water bath, hence this mixture
is employed for precipitation. .
Considerable quantities of alkali salts should be absent, for in
their presence .the precipit9-te is contaminated with zinc sodium
(or potassium) phosphate. These may be removed by double pre-
cipitation; the precipit.ate is dissolved in the minimum volume of
dilute hydrochloric acid and reprecipitated.
B. Zinc sulphide, znS, and subsequent weighing either as ZnS or
as ZnNH.P04 • The zinc is precipitated as zinc sulphide under
carefully controlled conditions. By the use of a chloroacetic acid-
sodium acetate buffer mixture, a pH of 2·6 to 2·7 is obtained, and
the sulpbide separates in a granular form. This method may be
used for the separation of zinc from manganese, nickel, cobalt (con-
centration ::I> 6 per cent of zinc). iron, a1uminium, chromium, and
uranium. The zinc sulphide may be weighed as such after heating'
with excess of pure sulphur in a current of hydrogen, or it may be
dissolved In 1: 1 hydrochloric acid, the hydrogen-sulphide remov~d,
and precipitated and weighed as in A.
• C. Zinc quinaldinate, Zn(C1oH60N2)2,H20. Quinaldinic acid or its
sodium salt precipitates zinc quantitatively from dilute acetic acid
or slightly ammoniacal solutions. Iron, aluminium, chromium,
beryllium, titanium, aI\d uranium interfere in acid 'solution, but in
tte presence of alkali tartrate'in'a~aline solution only zinc preci1?i-
tates. The reagent is expensive, but may be recovered (see Section
I,62N).
D. [Zn(C~H5N)J(CSN)2; pyridine method. This method is a very
rapid one, but unless the various wash solutions are carefully pre-
pared, low results will be obtained. The complex may be kept
unchanged in a vacuum desiccator 'for 2 to 3 hours (see Section
I,62L). Large quantities of ammonium salts must not be present as
these exert a slight solvent action upon the precipitate. If the
solution is strongly acid, it must be evaporated to dryness and the
residue dissolved in water.
E. Metallic Zinc. For details of the electrolytic determination
see Section IV, 84.
A. Determination of zinc as the ammonium phosphate or as
the pyrophosphate. Procedure. To the solution containing,
about 0·1 g. of Zn (1) add 5 g. of ammqnium chloride, 10 ml.
of 2N sodium acetate, and dilute to 1"40 ml. Heat on a water
bath, and, when hot, treat with 10 ml. of 10 per cent A.R.
ammonium phosphate (NH4)2HP04 added slowly from a pip-
ette or burette. Heat on' a water bath for 2 hours in a covered
beaker,. filter through .a weighed crucible (Gooch, sintered
glass or porous porcelain), wash with about 150 ml. of cold
water, 5 m!. of 95 per cent alcohol, and dry to constant weight
at 105°C (ca. 1 hour). Weigh as ZnNH 4P0 4.(2)
To weigh as the pyrophosphate, filter through a weighed
silica Gooch, porous porcelain, or quartz filter crucible
(Section n, 3~), and ignite slowly at 900°C in an electrically
heated muffle furnace to constant weight. If a muffle furnace
is not available, place the crucible within a larger nickel
crucible (Fig. 2-33), heat with a Me~er type of burner slowly at
first and then to redness until constant weight is attained.
Weigh as ZP2P207'
Note. 1. A suitable solution for practice may be prepared by dissolving
about 0·4 g., accurately weighed, of A.R. zinc sulphate in about 75 m!. of
water. A better method is to prepare pure zinc ammonium sulphate
ZnSO.,(NH.).SO"GH 2 0 by mixing equimolecular amounts of A.R. zinc
sulphate and A.R. ammonium sl11phate dissolved in boiling water, and
recrystallising the product twice from hot water. The crystals are air dried,
and 0·6 g., accurately weighed, is dissolved in 75 m!. of water.
2. A ~ore rapid, but less accurate method of weighing the precipitate,
collected on a sintered glass or porous porcelain crucible, is as follows (J. Dick,
1930). After washing with cold water, wash the zinc ammonium phosphate
5 or I) times with small volumes of Tectifi.e~ spirit and then 5 times with small
portions of anhydrous ether.. Suck the precipitate dry at the pump for 10
minutes, wipe the outside of the crucible dry with a clean I'lnen cloth, leave in a
vacuum desiccator for 10 minutes, and weigh as ZnNH.PO.,H,O which con-
tains 33 ·28 per cent of Zn.
B. Determination of zinc as sulphide. Procedure. The
zinc solution may contain up to 0·3 g. of Zn as' chloride or
nitrate, and should occupy a volume of 10 to 20 m!. (1).
Treat the solution dropwise with 2N sodium hydroxide until
there is a slight permanent turbidity. Add 10 ml. of 2N
chlor,oacetic acid (ClCH 2COOH) with stirring and, after the
solution has become clear, introduce 10 ml. of N sodium
acetate. Transfer quantitatively to a conical flask ~nd dilute
to 150 m!. Fit the flask with a two-hole rubber stopper
carrying an inlet tube extending well down into tb.e flask and
an outlet tube flush with the bottom of the stopper. Pass in
hydrogen sulphide rapidly for 10 to 15 minutes, and allow to
stand for 10 to 20 minutes when the white -and granular
precipitate of zinc sulphide will settle to the bottom of the
fla~k. Filter on a weighed porous porcelain crucible (A2)
and wash 4 times with solution 1 (2) if other metals are present,
followed by 4 times with solution 2 (3). When zinc alone is
present, only solution 2 is necessary. . .
Weighing as znS. Dry the cruci"9le at,.llO° to 120°C, cover
the precipitate with a layer of pure sulphur (which has been
. recrystallised from carbon bi-
sulphide), place a perforated Rose
crucible cover on the crucible
(Fig. 4-1), pass a stream q,fhydro-
gcn through it (the wash bottle
contains concentrated sulphuric
acid), and heat the crucible at
first over a small flame until
the excess of sulphur has dis-
~ appeared, and finally so that the
bottom of the crucible glows very
FIG. 4-1.
faintly for a few minutes. Allow
the crucible to cool in hydrogen,
",nd weigh as ZnS after leaving in a desiccator fdr 1;3 minutes.
Weighing as ZnNH 4P0 4 • Dissolve the precipitated zinc
sulphide in 25 to :H) 1]11. of 1 : 1 hydn1chloric acid, followed IJY a
'little water. Boil the s01ution to remove hydrogen sulphide,
and almost neutralise with ammonia solution. Complete the
estimation as in A. Weigh as ZnNH 4 P0 4 •
Notes. 1. If the original solution is acid, evaporate 'to dryness. and dissolve
in 10 to 20 m!. of w a t l ' r . · . •
2. Solzttioll 1: lUm!. of 2N chloracetic acid and 10 tnt. of N sodium acetotc
in 150 mL water and saturated with hydrogen sulphide in the col,\.
3. Solution 2: four per cent acetic acid saturated with hydrogen sulphide
in ~the cold.
C. Determination of zinc as quinaldinate. Procedure.
The solution (150 mt) should be neutral or weakly acid with
acetic acid, and contain about 0·1 g. of_Zn. Precipitate in the
cold by the slow addition, with stirring, of ~.3 to :30 1111. of the
reagent. (for preparation, see Section I, 62N). The precipitate
will settle out completely after about 20 minutes. Filter
through a weighed Gooch, sintered glass or porous porcelain
crucible,. wash ,,,ith alconol, and dry at 105 to llO°C to 0
constant weight. Weigh as Zn(C]oH 60 2 N)2,H zO.

Grav£metric A ttalysis 553
D. Determination of zinc by the pyridine method. Proce-
dure. The solution (7:1 mL) should conta.in auout 0·05 g. of
zinc (1) and be neutrn.l or very faintly acid. To the cold
solution add 1 g. of solid A.I{. ammonium thiocyanate,
followed by 1 m!. of pure pyridine. Shake vigorously, when
a white crystalline precipitate will separate. (Precipitation
may also be carried out in hot solution; the complex separat~s
in cOlPparatively large crystals on cooling.) Allow to stand
for 15 minutes, and stir frequently. Filter through a weighed
sintered glass or porous porcelain crucible, and transfer the
precipitate to the crucible with the aid of solution 1. Wash
the precipitate 4 times with solution 2, then wash the walls
of the crucible with 1 ml. portions of solution 3 (use a 1 or 2
m1. pipette for this process), and finally 1) to 6 times with
1 to 2 ml. volumes of s.olut£on 4. It is important to suck well
on the pump between each washing; it is also advantageous
to stir the precipitate with a thin glass rod when washing
with solutions 3 and 4. Dry the crucible and precipitate in a
vacuu m desiccator (preferably of tpe type shown in Fig.
2-10m for 15 minutes, and weigh. Repeat the drying process
until the weight is constant. Weigh as [Zn(CsHsN)2J(CNSk
Solution 1. 100 MI. of water containing 0·3 g. of NH 4CNS and
0·5 m1. of pyridihe, .
Sollttion 2. 85·5 MI. of water, 13 m!. of 95 per cent alcohol, 0·1 g.
of NH 4CNS,. and L·5 m1. of pyridine.
Solution ,j. 10 MI. of absolute alcohol + 1 m!. of pyridin<;.
Sollttion 1. 15:\1.1. of absolute (sodium dried) ether + i drops of
pyridine.
If the wash solutions have been prepared, the determination
should be completed withi,; an ho.ur.
Note. 1. For practice in this estimation, the student may employ about
V·25 g., accurately weighed, of A.R. zinc sulphate, or about 0·3 g. of pure
zinc ammonium sulphate (see Note to determination A).
IV, 35. MANGANESE

Discussion. The only method which is at all widely used for the
gravimetric estimation of manganese is the precipitation as mangan-
ese ammonium phosphate MnNH4P0 4,H20 in slightly ammoniacal
solution containing excess of ammonium salts. The precipitate may
be weighed in this form after drying at 100° t9 105°C, or it may be
ignited and subsequently weighed as maIfganese pyrophosphate
Mn2P 20 7• The latter procedure is by far the better pne. The
method is, however, tlf limited application because of the intl'rfering
influence of numerous other elements. Volumetric methods are
generally preferred (see Chapter III).
T*
554 Quantitative Inorganie Analysis
Determination of manganese as the ammonium phosphate
or as the pyrophosphate. Procedure. The solution (200 mi.)
should be slightly acid, contain not more than 0·2 g: of Mn in
200 mI., and no other cations except those of the alkali
metals (1). Almost neutralise the solution with dilute
ammonia solution, add 20 g. of ammonium chloride, and a
c<;msiderable excess of di!lmffionium hydrogen phosphate
(NH4)2HP04 (say, 2 g. of the solid). If a precipitate fOlJIls at
this point, dissolve it by the addition of a few drops of I : 3
hydrochloric acid. I-reat the solution almost to boiling (90 0
to '95°C), and add I : 3 ammonia solution dropwise and with

constant stirring until a precipitate (Mn a(P0 4):!') begins to
form; immediately suspend the addition of the alkali.
Continue the heating and stirring until the precipitate be-
comes crystalline (MnNH 4 PO"H 20). 1:hen add another drop
or two of ammonia solution, stir as before, etc., and. so con-
tinue until no more precipitate -is, produced and its silky
appearance remains unchanged. The precipitate must be
maintained at 90 0 to 95°C throughout; a large excess of
ammonia solution must be avoided. Allow the solution to
stand at room temperature (or, better, at O°C) for 2 hours.
Filter through a: quantitative filter paper or through a weighed
porous porcelain crucible, and wash the precipitate with cold,
10 per cent ammonium nitrate solution until free from
chloride. Dry at a gentle heat, ignite at as, lpw ~ tempera-
ture as possible until the carbon is oxidised (2), and then heat
at 1000°C (in an electric crucible furnace or within a larger
nickel crucible) to constant weight. Weigh as 'Mn2P 20 7 •
Alternatively, but less desirably, the precipitate in the porous
porcelain crucible may be dried at 1(~0° to 105°C to-constant
weight and weighed as MnNH 4 PO"H 2 0 (3); in this case, a
Gooch or sintered glass crucible may also be used.
Notes. 1. A suitable sorution for practice may be prepared by one of the
following methods :
(a) Dissolve 0'7 g., accnrately weighed, of A.R. manganQus sulphate
MnS0 4,4H 20 in 200 ml. of water.
(b) Dissolve 0·5 g., accurately weiglted, of it.R. potassium permanganate
in very dilute sulphuric acid, and reduG,e the solution with sulphur dioxide or
with alcohol (see reduction of potassium dichromate, Section IV, 30C). Re-
move the excess of sulphur dioxide or of acetaldehyde (and alcohol) by boiling.
Dilute to 200 m!.
2. These remarks apply, of course, when filter paper is used.
i. The p;ecipitate, collected on porous porcelain or.sintered glass crucible,
may be weighed more rapidly but less accurately as follows (J. Dick, 1930).
Wash the manganese ammonium phosphate 5 or 6 times with rectified spirit,
5 times with small volumes of anhydrous ether, and then suck the precipitate
Gravim.etric Analysis 555
dry at the pump for 10 minutes. Wipe the outside of the crucible dry with a
clean linen cloth and leave in a. vacuum desicca.tor for 10 minutes. Weigh
as MnNH.PO•• H 20 which contains 29.53 per cent of Mn.
IV,36. VANADIUM
Discussion. This element, as vanadate, may be determined in one
of the following forms :
A. Mercurous vanadate HgsVO, and subsequent weighing as
vanadium pentoxide V20 5• 'Phis method is applicable only in the
absence of substances which gives precipitates with mercurous
nitrate solution (e.g., chromate, arsenate, molybdate, tungstate and
phosphate). Actually the precipitate obtained in the determination
consists of a mixture of mercurous ortho- and pyro-vanadates; this
is ignited to, and weighed as, vanadium pentoxide. The solution
may contain free nitric acid but not ammbnium salts.
B. Silver vanadate, Ag SV0 4• Vanadates are .precipitated by
.excess of silver nitrate solution in the presence of sodium acetate,
followed by boiling, as silver orthovanadate. The following reac-
.tions occur with a solution of a metavanadate :
2NaVO a + 2Na.C2H:P2 + H 20 ~ Na,V20 7 + 2H.C2H s0 2 ;
Na"V 0 + 4AgNO a ~ Ag,V 0 + 4NaNO a ;
2 7 2 7
Ag"V2 0 7 + 2AgNOa + 2Na.C H a0
2 2
~ 2AgsVO" + 2H.e zH sOz + 2NaNO s.
Volumetric methods (see Chapter III) are, however, more con-
venient, less influenced by interfering elemenis, and are generally
preferred.
A. Determination of vanadium as mercurous vanadate and
ignition to vanadium pentoxide. Procedure. The alkali
vanadate solution (100 ml.) should be faintly acid with nitric
acid, .and contain not more than 0·1 g. of vanadium as
vanadate in 100 ml. of solution. Add 5 ml. of 20-volume
hydrogen peroxide and then mercurous nitrate solution (1)
dropwise until, after the precipitate has settled, a further
addition of 2 to 3 m!. of the reagent causes no precipitation.
Boil the covered beaker for 30 minutes to destroy the excess
of hydrogen peroxide, and then allow the grey precipitate to
settle. Filter-the precipitate through a Whatman No. 40 or
No. 540 filter paper, and wash it with water.containing a few
drops of mercurous nitrate solution. Dry the filter and
precipitate in the steam oven (or at 100°C), ash the filter paper
apart from the precipitate (Section n, 36e) in a weighed
porcelain, sil:ka or platinum crucible, ignite gently at first and
then strongly with a Meker type burner to constant weight.
Alternatively, the precipitate may be collected in a weighed
:')56 Quantitati'Ue I norga:nic Analysis
porous porcelain crucible; ignition is carried out by placing
the latter in a nickel crucible (Fig. 2-33) or in an electric muffie
furnace; Weigh as V20 5 . Owing to the poisonotts nature of
mercllry fumes, the whole of the operatioll concerned with the
ignition must be carried out in a fume cupboard which is pro-
vided with a good draught.
Note. 1. For preparation of mercurous nit7ate solution, see Note 1 in
Section IV, 30B.
B. Determination of vanadium as silver vanadate. Pro-
ced~tre. Neutralise the solution (200 mI.), containing not
more than 0·2 g. of alkali vanadate, if acid, by ilqueous sodium
hydroxide, or, if alkaline, by the addition of nitric acid to the
boiling soluJ:1on until it_becomes yellow, followed by decolouri-
sat ion with dilute ammonia solution. Add 3 g. of ammonium
acetate, 0·5 ml. of concentrated ammonia solution, and then
excess of silver nitrate solution, heat to boiling and then keep
on a steam bath foc' 30 minutes. Test for complete precipi-
tation with more silver. nitrate solution; if a turbidity is
produced, boil the liquid until it becomes clear. Allow the
dense brown pr~cipitate of silver vanadate to settle, and
collect it on a weighed filter crucible (Gooch, sintered glass or
porous porcelain), wash with hot water, dry at lloac, and
ignite gently to constant weight within a larger crucible.
Weigh as Aga V0 4.
1V,37. URANIUM
Discussion. Uranium, as uranyl salts, may be determined in either
of the following forms:
A. Ammonium. diuranate, (NH4)2U207' and ignition to uranous
uranate, UaOs. The uranyl solution is precipitated by ammonium ..
hydroxide solution as ammonium diuranatc; theJatter is ignited to,
and weighed as, uranium oxide U aOs, All elements that arc preci-
pitable by ammonia solution must be absent. The reagents and
solutions must be free from carbonates and organic matter, both of
which prevent precipitation of the uranium.
B. Uranyl" oxinate," U02(C9H60N)2,CgH70N. -1'he formula of
the compound is noteworthy, for it differs from all other metallic
" oxinates" (compare Section I, 620). This method may also be
employed for the volumetric determination of uranium with stan-
dard potassium bromate solution (compare Section m, 138). Here
1m!. N KBrOa ~ O'Ou)85 g. U.
A. Determination of uranium as ammonium diuranate.
Procedure. The uranium should" be in the form of uranyl
sulphate in~ilute sulphuric acid solution; the uranyl sulphate
content should not exceed 1 per cent. Add a few lirops of
methyl red indicator, heat to boiling, and treat with dilute
ammonia solution that is free from carbonate until the indica-'
tor just becomes distinctly yellow (the precipitate also is
.yellow). Add a W,hatman a:ccelerator or one-third of an
ash less tablet, and warm for f or 2 minutes. Filter off on a
Whatman No. 41 or No. 541 filter paper, and wash the precipi-
tate well with a hot 2 per cent solution of ammonium nitrate.
Dry the wet paper and precipitate in a platinum or a Main-
Smith crucible at as Iowa temperature as possible until the
carbon is destroyed, and then heat strongly over i Mcker (or
similar) burner, the crucible being placed in a slan.ting position
so as to maint?-in good oxidising conditions (witb the :Main-
Smith crucible, the lid should be removed and placed on one
side). Cool in a desiccator and weigh. Repeat the ignition
until constant weight is attained. Weigh as U 308'
B. Determmation o.f uranium with oxine. Procedure.
The uranium should be in the form of uranyl nitrate or
chloride in 1 to 2 per cent acetic acid solution (1), and the
solution may contain up to 0·3 g. of U. Add 5 g. of ammo-
nium acetate, heat to boiling, and add slowly a four-fold excess
of tIre oxine reagent (2):" Digest on the water bath for a
short time, and allow to cool. When .cold, filter on a weighed
filter crucible (Gooch, sintered glass or. porous porcelain),
wash thoroughly with hot water untiT the excess of oxine has
been removed (ferric chloride test), and then several times
with cold water. Dry at 105° to 110°C to constant weight,
Weigh as UOZ(CUH60N)z,C9H70N.
Notes. 1. If the solution contains mineral acid, almost neutralise with
ammonia solution, add 5 g. of ammonium. acetate and then sufficient acetic
acid to give a 1 to 2 per cent acid solution.
2. The oxine reagent is prepared as follows. Dissolve 3 g. of oxine in the
minimum volume of glacial acetic acid, dilute to 100 lill. with water, add dilute
ammonia solution dropwise until a permanent slight precipitate separates, and
just disstllve the latter by the addition of a little dilute acetic acid. Alterna-
tively; th~ method of Section I, 62P may be used.
IV, 38. THORIUM

Discussion. ThoriuI!l may be determined in either of the follow-
ing forms:"
A. Thorium oxalate, Th(C20 412, and subseq'uent ignition to thoria.
ThO z. This method is of rather limited application since phosphates
and all the rare earths must br absent. Thorium oxalate is soluble
in ammonium oxalate solution and in strong mineral acids (concen-
tration greater than, for example, O'4N hydrochloric acid).
B. Thorium sebacate, Th(C4H sC0 2)z, and subsequent ignition to
thoria, Th0 2• This procedure permits of the separation by a single
. precipitation of thorium from relatively large amounts of the rare
earths (Ce, La, Pr, Nd, Sa, Gd) and also from quadrivalent ·cerium.
A. Determination of thorium as oxalate. Procedure. The
thorium should be in hydrochloric a-cid solution (100 mI.)
containing not more than 3·5 per cent of hydrochloric acid by
volume and not more than 0·1 g, of T1;l. Heat the solution
(100 ml.) to bqiling, and add slowly and with constant
stirring sufficient of a boiling Ie per cent solution of oxalic
acid to combine with- all the thorium and leave an excess of
20 m!. AllQ,w the solution to cool and stand overnight. Filter
through a quantitativ~ filter paper, wash with a solution
containing 3·5 mI. of concentrated hydrochloric acid and 2·5
g. of oxalic acid per 100 mI., and ignite with,the full heat of
a Meker (or similar) burner (ca. 1l00oq in a weighed platinum,
porcelain or silica crucible to constant weight. Weigh as
Th0 2 •
B. Determination of thorium as sebacate. Procedure.
The solution (100 ml.) should be neutral or faintly acid, and
contain not more thanO·lg. of Th. :tIeat the solution to boiling
and add slowly and with constanf stirring a h'ot almost
sa_turated solution of pure sebacic acid in sli~ht exc~ss.
The precipitate is' voluminous, but granular, and therefore
easily manipulated. Filter off immediately, wash thoroughly
with hot water, dry, ignite, and weigh as Th0 2 (see A).
IV, 39. CERIUM·

Discussion. Cerium may be determined as cerie iodate, Ce(IO a)4'
which is ignited to, and weighed as, eerie oxide.,Ce0 2 • Thorium (and
zirconium) must, however, be- first removed (see Section IV, 38) ;
the method is then .applicable in the presence of relatively. large
quantities of rare earths. Volumetric methods (see Section m, 82)
are generally preferred.
Determination of cerium as eerie iodate and subsequent
ignition to ceria. Procedure. The solution should not exceed
50 m!. in volume, all metallic elements. should be present as
nitrates, and the cerium content should not exceed 0·11 g.
Treat the solution with half its voJume of concentrated nitric
acid, and add 0·5 g. of potassium bromate (to oxidise the
cerium). When the latter has dissolved,. add 10 to 15 times
the theoretical quantity of potassium iodate in nitric acid
solution (1) slo:vly and with constant stirring, and allow the
precipitated eerie iodate to settle. When cold, filter the
precipitate through a fine filter paper (e.g., Whatman No. 42
or No. 542), allow to drain, rinse once and then wash back
into the beaker in which precipitation took place by means of a
solution containing 0·8 g .. of potassium iodate and 5 ml. of
concentrated nitric acid in 100 ml. Mix thoroughly, collect
the precipitate on the saJ)1e paper, drain, wash back into the
beaker with hot water, boil, and treat at once with concerr-
trated nitric acid drop wise until the precipitate just dissolves
(20 t9 25 ml. of acid are required per 0·1 g. of ceria). Add
0·25 g. of potassium bromate and as much potassium iodate-
nitric acid solution as before. When cold, collect the eerie
iodate upon the same filter paper, wash once with the washing
solution, return to the beaker, stir with the washing solution,
filter again, and wash thrice with the same solution. Place
the filter paper and precipitate in the same beaker, add [) to
8 g. of oxalic acid and 50 m!. of water, and heat to boiling.
After all the iodine has be,en expelled, set aside for several
hours, filter, wash with cold water, dry and ignite to constant
weight in a platinum crucible. WeIgh as CC02'
Note. 1. This is prepared by dissolving 50 g. of potassium iodate in 1(\7 ml.
of concentrated nitric acid,"and diluting to 500 m!.
IV,40. TITANIUM
Discussion. Titanium may be determined in one of the following
forms:
A. Titanium "oxinate," TiO(CgHsON)2' Full C'xperimental
details of this method are given in Secti<?n m, 138, Procedure F.
Alkaline earth metals, magnesium and the alkali metals do not
interfere. The precipitate of TiO(C gH sON)2,2H 20 is collected in a
weighed sintered glass or porous porcelain crucible, dried to cons,ant
weight at llOac, and weighed as TiO(C 9 H 60N)2'
B. Titanium dioxide. Ti0 2• after precipitation with tannin and
antipyrine. This me"thod affords ~ separation from iron, aluminium,
chromium, manganese, nickel, cobalt and zinc; and is applicable in
tI,e presence of phosphates and silicates. Small quantities of titan-
ium (2 to 50 mg.) may be readily estimated.
2. Titanium dioxide, Ti0 2• after precipitation with selenious acid.
This procedure may be applied in the presence of aluminium, beryl-
lium, manganese, nickel, cobalt, the alkaline earth metals arid
magnesium; zinc is not precipitated, but may be occluded if the
selenious acid is added too rapidly.
D. Titanium dioxide, Ti0 2, after precipitation with para-hydroxy-
phenylarsonic acid. This procedure will separate titanium from
most other commonly occurring ions by a single precipitation,
Zirconium, tin, eeric cerium and hydrogen peroxide must be absent.
B. Determination of titanium with tannin and antipyrine.
Procedure. The Ti content of the solution should not exceed
0·1 g. of TiO z, and the titanium should be present as the
sulphate or chloride. Add dilute ammonia to the solution
llntil the odour persists, then (cautiously) 10 ml. of concen-
trated sulphuric acid, and 40 tnI. of 10 per cent tannin solu-
tion. Di~ute' to 400 ml., stir thoroughly, and cool. Intro-
duce a 20 per cent aClueous solution of antipyrine (" phena-
zone") with constant stirring until an orange-red flocculent
precipitate is obtained. Stop the stirring, and continue the
addition of the antipyrine solution until a white cheese-like
precipitate (produced by the interaction of tannin and anti-
pyrine) is formed in addition to the red precipitate. Boil the
mixture, remove the flame, add 40 g. 'of ammonium sulphate,
and allow to cool with occasional stirring. Filter the bulky
precipitate through a What man No. 41.or No. 1)41 filte~ paper,
supported on a Whatman filter cone (hardened, No. 51), with
slight suction, and wash with a solution composed of 100 ruI. of
water, 3 ml. of concentr.ated sulphuric acid, 19 g. of ammo-
nium sulphate and 1 g. of antipyrine. Dry the precip~tate at
100°C, transfer to a weighed crucible, heat gently at first and
then ignite.to constant weight. Weigh as 1'10 2 ,
Note. If the wet precipitate is heated directly. caking occurs which renders
the complete oxidation of the carbonaCeO!lS matter very slow. If alkali
metu.ls were originally present. the ignited titania must be washe(l with hot
water, filtered. and re·ignited to constant weight.
C. Determination of. titanium with selenious acid.
Procedl~re. The titanium should be in solution in dilute acid,
preferably- hydrochloric acid in which precipitations an.{i
separations are mQre facile than in dilute sulphuric acid;
the solution.should contain not more than about 0·1 g. of Ti.
Treat the cold acitl solution of titanium. with dilute ammonia
solutij)!l until the acid concenfration is reduced below 0·2N,
and then add with constant stirring [j per cent aqueous
sclenious acid solution until the precipitate coagulates. The
precipitate has the composition H 2 [Ti0 2 ,SeO aJ,2H 20 and
hccomes anhydrous at 1l0°C; it adsorbs traces.of selenious
acid, however, so that weighing in this form is not permissible.
Add a Whatman accelerator (or one-third of an ashless tablet)
and allow the precipitate to settle for a few hours. Filter
the precipitate through a Whatman No. 41 or No. 541 filter
paper, supported upon a Whatman filter cone (hardened, No,
iiI), with slight suction, and wash with cold water. Transfer
the wet precipitate to a weighed porcelain, silica or Main-
Smith crucible, dry over a small flame, heat until all the
selenium is expelled, and ignite to constant weight. Weigh
as TiO z.
D. Determination of titanium with para-hydroxy-pheny!-
arsonic acid. Procedure. Dissolve the sample (1) containing
not more than about 0·06 g. of TiO z in sulphuric or hydro-
chloric acid, and dilute to 200 ml. The amount of the acid
present should be such that the solution will be approximately
but not more than O'60N in l\ydrochloric acid or I·S0N in
sulphuric acid after the reagents have been added and the
precipitation is complete. Heat .the solution to boiling; if
iron is 'present, add 2 to 3 g. of A.R. ammonium thiocyanate:
add 100 m!. of a 4 per cent aqueous solution of para-Jtydroxy-
phenylarsonic acid (2). Boil gently for 15 minuteS' to coagu-
Jate the precipitate. Allow to cool to room tempe~ature, and
filter with suction on a Whatman No. 542 or No. 42 filter
paper supported on a filter cone (Whatman, No. 51, hardened).
Wash the precipitate 5 or 6 times with a wash liquid of
0'2:')N hydrochloric Qr sulphuric acid containing about 0·5 g.
of the solid reagent per 100 m!. (if iron is present, 1 to 2 g.
of ammonium thiocyanate should also be added to each 100
ml. of -wash liquor. Finally wash tha precipitate 2 or 3
times with 2 per cent aqueous ammonium nitrate solution.
Transfer the filter to a Main-Smith crucible, ignite gently at
11rst until all the,carbon is burnt off (thi~ operation must be
carried out in a fume chamber provided with an efficie1f..t
draught) and then with the full blast of a Fisher burner until
constant weight is attained:- Weigh as Ti0 2 •
Notes. 1. For practice in this determination the U.S. Bureau of Standard
.. Iron Ore. No. 29" may be used (see Section A, 8). Dissolve 4 g. of-this in
100 ml. of dilute hydrochloric acid, and filter. Fuse the undissolved residue
with sodium carbonate, wash the melt into the I!Iain filtrate, remove the silica
in the ub'Ual manner, add 4 g. of A.R. ammonium thiocyanate, dilute to 200 to
250 mI., and continue the estimation as above.
2. A.R. para-hydroxy-phenylarsonic acid is supplied by the MaIIinckrodt
Chemical Works, New York, U.S.A.; see Section A,9.
IV, 41. ZIRCONIUM

Diswssion. Zirconium may be determined in one of the following
forms:
A. Secondacy zirconium phosphate, ZrHZ(P0 4)2' and subsequent
ignition to the'pyrophosphate, Zr2P207' Zirconium is quantitatively
precipitated from solutions containing as much as 10 per cent by
Quantitative Inorganic Analysis
volume of sulphuric acid by excess of diammonium hydrogen phos-
phate provided that the solution is allowed to stand at 40° to 50°C
for a few hours. The precipitate is washed with a moderate amount
of 5 per cent ammonium nitrate solution to remove excess of phos-
phate; prolonged washing results in hydrolysis of the zirconium
phosphate and consequent loss of phosphate, thus introducing
errors which may be as high as 1 or 2 per cent. If titanium is
present, excess of hydrogen peroxide must be added before the
addition of the phosphate precipitant. Cerium should preferably
be in the teryalent condition; a double precipitation is required in
the presence of thorium. ,-
B. Basic zirconium selenite an<\subsequent ignition to the dioxide,
Zr0 2• The zirconium is precipitated as the basic selenite with selen-
iollS acid in hot dilute hydrochloric acid solution, the precipitate
washed with dilute hydrochloric acid, and then ignited to, and
weighed as, Zr0 2 • No other acids should be present, and the hydro-
chloric acid content should be 5 per cent and not over 7 per cent by
volume. This method enables a separation to be effected by a single
precipitation from the rare earths (cerium being in ~he tervalent
condition) and from aluminium. Iron, if present up to 10 per cent
of the weight of the zirconium, does not interfere if precipitation is
made in dilute solution. (100 to 200 m!. of solution containing 0·05
g. Zr0 2) and double the quantity of seleniOlls acid solution is used
for preCipitation.
C. Zirconium n-propylarsonate and subsequent ignition to the
dioxide. zr0 2 • Zirconium is quantitatively precipitated in dilute
suJphvric acid solution by aqueous n-propylarsonic acid; tin
interferes, but may be removed from the zirconium oxide residue
obtained upon ignitwn by adding powdered ammonium iodide, and
re-igniting (see Section IV, 22A). Titanium, aluminium, chromium,
cobalt, nickel, copper, uranium, vanadium, thorium, molybdenum
and tungsten do not interfere. Complete precipitation of the.
zirconium takes place in the presence of large quantities of iron, but
a: small quantity of iron is adsorbed upon the zirconium precipitate;
the iron is completely removed by reprecipitation. The method is
therefore of particular value for the analysis of complex steels con-
taining zirconium. FuU particulars are given in Section I\', 92L ;
the method for the determination of zirconium in a simple compound
or mixture should be obvious from these details.
D. Zirconium para-hydroxy-phenylarsonate and subsequent
ignition to the dioxide. The procedure is similar to that already"
given for titanium (Section IV, 40D) except that a mineral acid
concentration of 2·5 to 3'ON is employed, and somewhat less reagent
is necessary. Phosphates must, of course, be absent. The method
is particularly applicable to the determination of zirconium in the
presence of a large amount of iron.
A. Determination of zirconium as pyrophosphate. Procedure.
The zirconium solution (100 mI.) should contain about 10 per
cent by volume of concentrated sulphuric acid and about
0·05 g. of Zr. Add a few ml. of 100-volume hydrogen peroxide
(" perhydr.ol ") to oxidise any titanium present and keep it in
solution; if titanium is known to be absent, the addition of
. hydrogen peroxide is unnecessary. Add a freshly prepared
10 per cent aqueous (NH 4)zHP0 4 soluti,on in 10 to 20 fold
excess over that required to form ZrH Z(P0 4)z' Adjust the
acidity to 10 per cent by volume of sulphuric acid and digest
at 40° to 50°C for a few hours; it is advisable to add a
Whatman accelerator to assist in the subsequent filtration.
Allow the -precipitate to settle (if only a little Zr is present, it
is best to leave it overnight), filter through a quantitative
filter paper, and wash w1th cold 5 per cent arp.monium nitrate
solution until free from phosphate (not more than 200 m!.
should be used). Transfer the wet paper and precipitate to a
weighed porcelain, platinum or, best, Main-Smith crucible,
heat very gently until all the paper is charred, ignite very
carefully until all the carbon has been oxidised, and then
ignite to constant weight .. Weigh as Zr ZP Z0 7 •
B. Determination -of zirconium with s~enious acid. Proce-
dure. The solution (200 mI.) should contain about 5 per cent
by volume of hydrochloric acid (sulphuric acid is undesirable)
and not more than 0·2 g. of zirconia. Treat with 20 ml. of
12·5 per cent aqueous selenious acid solution, and boil for a
few'minutes. Allow the precipitate of basic selenite to settle,
filter through a quantitative filter paper, wash with hot. 3 per
cent hydrochloric acid containing a little selenious acid, dry
in a weighed porcelain, platinum or Main-Smith crucible, and
ignite to constant weignt. Weigh as Zr0 2 •
IV, 42. THALLIUM

Discussion. Thallium may be determined in either of the foll~w
ing forms:
A. Thallous chromate, TliCr0 4• The thallium must be present in
the thalIo~s stat_e. If present as a thallic salt, reduction must be
effected with sulphur dioxide before precipitation; the excess of
sulphur dioxide is boiled off.
B. Thallous cobaltinitrite, Tl 3 [Co(N02 )]. It is possible to preci-
pitate a thallous salt quantitatively by means of a special sodium
cobaltinitrit~ reagent at 30° to 40°C in the presence of 5 per cent of
formic acid,.
564 Qlta",;titative Inorganic Analysis
A. Determination of thallium as chromate. Procedure.
The solution (100 mI.) should contain about 0·1 g. of T1, no
excessive amounts of ammonium salts, a~d no substances that
form precipitates with ammonia solution, or reduce potassilnn
chromate, or react with potassium or thallous chromate in
ammoniacal solution. Neutralise the thallium solution (lOO
mI.) with ammonia solution, and add 3 m!. in excess. "Heat
to about 80°C, and add 2 g. of potassium chromate in the form
of a 10 per cent so)ution slowly and with constant stirring.
Allow to stand at the laborJ.tory temperature for at least 12
hours. Filter through a weighed filter" crucible (Gooch,
sintered glass or porous porcelain), wash with 1 per cent
potassium chromate solution, then with 50 per cent alcohol,
and" dry at 120°C to constant weight. Weigh as T1 2Cr0 4 •
13. Determination ot thallium as cobaltinitrite. Procedure.
The solution (100 mI.) should contain about 5 per cent of free
formic acid, and about 0·1 to 0·2 g. of Tl; potassium and
other elements which give precipitates with the .sodi.um
cobaltinitrite reagent should be absent. Warm the solution
to 30° to 40°C and add an excess. of the reagent (1) at the
-same temperature. Allow to sfand" for 1 hour, collect the
scarlet precipitate \on a weighed sintcrcd glass or porous
porcelain crucible, wash with cold water, a-nd dry at lOoDe
to constant weight. Weigh as T1 3[Co(N0 2)6]' I
Note. 1. The reagent is prepared by mixing c<Jual volumes of (a) 28·6 g.
of cobalt nitrite and 50 mk of 50 per cent formic acid in 500 m!. of water. and
(b) 180 g. of sodium nitrite per 500 m!.
lV,43. CALCIUM
Discussion. Calcium may be tietE;;rmined in one of the following
forms :\ calci:nn, oxalate CaC20 4,H20. calcium carbonate CaC03, and
calcium oxide CaO. These methods are fully discussed in Section
IV, 10. A clean separation from magnesium with a single precipita-
tion may be obtained by precipitation as calcium molybdate CaMo04
(see Section IV. 94H and Section IV, 18A).
IV, 44. STRONTIUM
Discu:ssion. .Strontium may be estimated in one of the following
forms: .
A. Strontium sulphate. SrS04 • In this determination, the most
accurate for'strontium, calcium, barium and lead must be absent,
and the solution (preferably hydrochloric) should be as neutral as
possible. If considerable acid is present, this must be removed by
evaporation. Strontium sulphate dissolves appreciably in an acid
medium becaus~ of the reaction:
5rS0 4 + H+ ~ HS0 4- + 5r++.
Strontium sulphate has a solubility of about 0·14 g. per litre at the
ordinary temperature; the solubility is decreased by the addition
of a slight excess of sulphuric 'acid, and of alcohol.
B. -Strontium oxide, SrO. The strontium is precipitated as the
oxalate SrC 20 4 ,H 20, ann is ignited to, and \veighed a~, the oxide
SrO. The experimental details arc similar to those given in Section
IV.., 10 except that some -alcohol should be added to reduce the
solubility. Calcium, barium, magnesium and other interfering
elements must be absent.
A. Determination of strontium as sulphate. Procedure.
The solution (100 ml.) should contain about 0·2 g. of.Sr,and
be slightly acid with hydrochloric acid- (1). Add slowly a
ten-fold excess of dilute sulphuric acid, followed by a volume
of alcohol equal to that of the solution. Stir well, and allow
to stanel for at least 12 hours. Transfer the precipitate to a
weighed silica Gooch or porous porcelain crucible, wash with
7 [) per cent alcohol to which a few drops of sulphuric acid have
been added, and finally with pure alcohol (or rectified spirit)
until the washings are free from sulphate.' Dry and ignite
(in a larger nickel crucible) at dull redness to constant weight.
Alternatively, a filter paper may be used: here trye paper
should be burnt apart from the precipitate (Sectian II, a6C) (to
prevent possible reduction to the sulphide), and the residue
then ignited together with the main precipitate in a weighed
porcelain., silica or platinum crucible. Weigh as SrS0 4 •
Note. I •.A soLution for practice in this determination may be prepared by
dissolving u·:! to U·4 g., accurately weighed; of pure strontium carbonate
(the A.R. product may be obtained from the Mallinckrodt Chemical. Works,
U.S.A. ; see Section A, 9) in a littLe dilute hydrochloric acid (sec Section IV, 10).
and diluting to 100 ml.
IV, 45. BARIUM

Discllssion. Barium "may be determined in one of the following
forms:
A. Barium sulphate, BaS04• This method is most widely em-
ployed. The effect of various interfering elements and radicals
(e.g., calcimn, strontium, lead, nitrate, etc., which contaminate..the
precipitate) has been fully dealt with in Section IV, 6.
B. Barium chromate, BaCr04• This method is of limited appli-
cation because of the influence of numerous interfering elements. It
is USefill, however, in the separation of barium from both calciul)l.
and strontium. Barium chromate may be heated to a red heat
without decomposition. Ii green spots of chromic oxide, due to the
reduction of the chromate, should appear, continued ignition with
free access of air will result in re-conversion il1to chromate.
A. Determination of bar,ium as sulphate. Procedure. The
barium solution (100 ml.) should contain not more than 0·15
g. of barium (1), and not more tha.n 1. per cent by volume of
concentrated hydrochloric acid. Heat to boiling, add a
slight excess of hot N sulphl,lric acid slowly and with constant
stirring. Digest on the steam bath until the precipitate has
settled, filter:wash with hot water containing a few drops'of
sulphuric acid, and then with a little water until the acid is
removed. Full experimental details of the filtration, washing
and ignition processes are given in Section IV, 6. Weigh as
BaS04 •
about O·~ g., accurately weighed, of A.R. barium chloride in 100 mI. of water
and adding I ml. of concentrated hydrochloric acid.
B. Determination of barium as chromate. Procedure 1.
The solution (200 m!.) (1) should contain not more than 0·4
g. of Ba, and be neutral in reaction. Add about 1 m!. of 6N
acetic acid to the neutral solution (200 ml.), heat to boiling,
and treat with a slight excess of a ·hot dilllte solution of
'ammonium chromate (2) drop wise from a burette and with
constant stirring. Place the beaker on a water bath until the
'precipitate settles; test for c:;ompleteness of precipitation by
adding a little more of the reagent. Allow to cool, filter
through a weigl!ed Gooch or porous porcelain crucible, wash
with hot water until 1 ml. of the washings gives scarcely any
reddish-brown colouration with neutral silver nitrat~ solution.
Place the crucible inside a larger porcelain or nickel crucible,
and. heat to a temperature not exceeding dull-red heat u,ntil
the precipitate has a uniform bright yellow colour, and is
,constant in weight. Weigh as BaCr0 4 •
Notes. 1. A suitable solution for practice in this determination may be
prepared by dissolving about 0·4 g., accurately weighed, of A.R. barium chloride
in 200 m!. of water.
2. The ammonium chromate solution is prepared by dissolving 10 g. of
pure ammonium chromate (free from sulphate) in 100 m!. of water, and adding
dilute ammonia solution until the colour of the solution is clear yellow.
ProCedHre Z. The solution (100 ml.) should contain not more
than 0·30 g. of Ba, and be neutral in reaction. Add 10 m!.
ofN acetic acid and 5 g. of pure ammonium chloride, and stir
until the solid has dissolved. Heat to boiling, and add 10
ml. of 10 per cent A.R. potassium chromate solution slowly
ang. with constant stirring. Boil gently for 2 to 3 minutes,
and allow to stand overnight. Collect the barium chromate
on a weighed filter crucible (Gooch, porous porcelain or
sintered glass), wash with 50 ml. of cold water, and dry to
constant weight at 130°C. Weigh as BaCr0 4 •
IV, 46. MAGNESmM
Discussion. Magnesium may be estimated in one of the following
forms: A. Magnesium pyrophosphate, Mg2P207' This determination
is dealt with fully in Section IV,ll.
B. Magnesium "oxinate " (8-hydro:xyquinolate), Mg(CsH60N)2'
~O. The experimental details, applicable if only alkali metals
are present with the magnesium, have been described in Section
m,.l38, Procedure B.
For practice in this determination, the student may employ
about -0·2 g., accurately weighed, of A.R. magnesium sulphate
dissolved in 150ml. of water. The precipitated magnesium "oxinate"
is collected on a weighed sintered glass or porous porcelain crucible,
well washed with warm I % ammonia solution, dried at 100° to
105°C, and weighed as Mg(C gH 6 0N)2,2H 20.. Alternatively, the
precipitate may be dried at 130° to 140°.(;, and weighed as
the anhydrous compound Mg(C oH 60N)2'
Heavy metals, other than copper, cadmium and zinc, are pre-
vented from' interfering if the magnesium is precipitated from a
solution (originally neutral) to which 10 to 20 ml. of 2N sodium
hydroxide and 5 g. of sodium tartrate have been added. If copper,
cadmium and/or zinc are present they may be precipitated from
acetic acid solution by oxilie, and after their separation, magnesium
is determined in the filtrate by the usual method. If aluminium is
present, it is precipitated in neutral or weakly acid solution by oxine
(see Section m, 138, Procedure A) and the magnesium determined in
the filtrate.
In all cases where metals other than those of the alkalis occur in
solution, reprecipitation is necessary. The first precipitate of
magnesium oxinate is dissolved in the minimum volume of dilute
hydrochloric acid, 1 to 2 g. of ammonium acetate and a few drops
of the oxine solution added, and the solution made just alkaline
with dilute ammonia solution. The' mixture is boiled for a few
minutes to render the precipitate crystalline, and is then collected
and weighed...
IV, 47. SODIUM
Discussion. Sodium may be determined in one of the following
form§ :
A. Sodium sulphate, Na2S0 4• Any sodium compound of a volatile
acid may be copverted into sodium sulphate by repeated evaporation
with sulphuric acid. Some bisulphate NaHS0 4 is formed in the
process, and this is converted (via the pyrosulphate Na 2Sp7) into
the norIllal salt with some difficulty. The latter change is facilitated
by the addition of a little powdered ammonium carbonate; this is
because ammonium sulphate, which is completely volatilised on
heating, is formed:
Na 2S20 7 + (NH4)2COa = Na2S04 + (NH,hSO, + CO 2,
56R Quantitative Inorganic Analysis
B. Sodium zinc uranyl acetate, NaZn(U0 2) 3(C2H 30 2) 9,6H 20.
Treatment oI a concentrated solution of a sodium salt with a large
excess of zinc uranyl acetate reagent results in the precipitation of
sodium zinc uranyl acetate. This substance is moderately soluble
in water (58·5 g. per litre at 21°C) so that a special procedure must "be
adopted in the washing and collecting of the precipitate. Potas-
sium does not interfere provided not more than 0·05 g. is present per
m!. of solution immediately prior to the precipitation. Ammonium,
calcium, strontium, barium, magnesium, iron, aluminium, and
chromium do not interfere if present in reasonable amounts, but
lithium, which also forms a sparingly soluble triple acetate, must be
absent. Oxalates, tartrates, phosphates and arsenatcs, which yield
precipitates with the uranyl ion, should not be present. Sulphate
must be absent when potassium is present, for potassium sulphate
is sparingly soluble in-the reagent; sulphate may be converted into
chloride by treatment with barium chloride.
C. Sodium magnesium uranyl acetate, NaMg(U02)3(C~a02)~'
6·5H 20. This determination is similar to B except that magnesium
uranyl acetate reagent is used. The method is applicable to solu-
tions containing up to 25 mg. of Na, although with care 50 mg. may
'be estimated; the volume of the solution should be 5 ml. Ammon-
ium and calcium ions are practica}ly without effect; the maximum
quantities of potassium permissible with 50 ml. and lOO ml. of the
reagent are 0·1 g. and 0·25 g. respectively. The procedure may be
applied to the determination of sodium in felspar (see Section
IV,95F).
A. Determination of sodium as sulphate. Procedure. This
estimation is most conveniently carried Qut in a Main-Smith
crucible (Section II, 36B). Evaporate the solution (1) to
dryness in a weighed Main-Smith crucible on the water bath.
Transfer to a triangle (or to hot plate, the tenlperature of
which can be controlled), add a few ml. of concentrated
sulphuric acid dropwise on to the cover of the crucible, and
evaporate gehtly to dryness in the fume wPboartMllltil,fuming
ceases. Repeat this operation twice. Allow to ~ool, add a
few small pieces (about the size of a pea) of solid ammonium
carbonate to decompose any pyrosulphate present, and heat
to dull redness for 15 minutes. Allow to cool in a desiceator
for 25 minutes, and. weigh immediately at the end of this
period. Repeat th~ ignition until constant weight_is attained.
Weigh as Na 2S0 4 •
Note. 1. oA suitable solution for practice m<ty be prepared by weighing Ollt
accurately about 0·5 g. of A.H.. s()dium chlori'ue, and dissolving it in a little
water.
B. Determination of sodium as sodium zinc uranyl acetate.
Procedure. The final solulion !pay contain up to 0·05 g. of
Na per mI., but the method is best applied to solutions con-
taining not more than 0·008 g. of Na per m!. Treat the con-
centrated sodium solution (say, 1·5 m!.) (1) with 15 m!. of
zinc uranyl acetate reagent (2), and stir vigorously. Allow to
stahd for 1 hour, and filter through a weighed sintet:ed glass
crucible (G4). Wash the predpitate 4 times with 2 m!. por-
tions of the precipitating reagent (allow the wash liquid to
drain co_mpletely before adding the next portion), then 5 times
with 95 per cent alcohol saturated with sodium zinc uranyl
acetate (2 ml. portions), and fi'nally with a little dry ether
or acetone. Dry for 30 minutes only at 10;)oC. Weigh as
NaZn(U0 2) 3(C 2 H;P2)9,6H 20.
Notes. I. A suitable "olution for practice may be prepared by Iwaporating
20 m!. of O'IN ~o(lium chloride, prepared from the A.R. salt, to 1'5 m!. on a
water bllth.
:l. The reagent is prepared by mixing "qual vollhnes of solutions .4 and B
a illl filtering.
SO/Iltialz A : dissolve 20 g. of crystallised uranyl acetate UO.(C.H ,O.) •. 2H.O
in 4 ml. oj glacial acetic acid and 100 ml. of water (warming may be
necessary). . •
"solution B: dissolve 60 g. of crystallised zinc acetate Zn(C.H 30.). 3IT.O
in;) m!. of glacial acetic acid and lOll m!. of water.
c. Determination of sodium as sodium inagnesium uranyl

acetate. Procedure. The solution of the sodium salt 'should
be evaporated to !) m!. if 10 mg. or more of Na are present;
if 1 mg. is present', the volume' should be reduced to 1 to 2 m!.
Treat the solution (1) with excess of the clear reagent (2)-
as a rough guide the number of ml. of the reagent sl10uld be 10
times the number of mg. of Na present; excess of the reagent
does no harm. Immerse the vessel in a bath maintained at
20° ± 1°C, and stir vigorously (preferably by means of a
mechanical stirrer) for 30 to 45 minutes. Filter the precipi-
tate immediately through a weighed sintered glass crucible.
using gentle suction,' and after all the solution has run
through, wash the precipitate 4 to 5 times witll successive 5
'ml. portions of 91') per cant alcohol saturated with sodium
magnesium uranyl acetate. Dry for 30 minutes at 105°C,
and weigh as NaMg(U02)3(C2H302)9,6·5H20.
Notes, 1. A suitable solution for practice in the determination may contain
26 mg. of Nain 5 m!. -
2. The reagent is prepared as follC?ws :
Solution A : 21·3 g. of crystallised uranyl acetate, 15 g. of glacial acetic acid
and 250 Ill!. of water.
Solution B: 125 g. of crystallised magbesium acetate Mg(C zH.O')2,6H J O,
15 g. of glacial acetic acid, and 250 m!. of water .
Heat solutions A and B separately to 7Q·C until all the salts are dissolved,
and then mix the solutions at 70·C, and allow to cool to 20·C. Place the
vessel containing the mixed reagent in water at 20· ± 1°C, and maintain at
this temperature for 1 to 2 hours until the slight excess of salts has crystallised
out. Filter through a dry filter into a dry bottle. The solution..is stable for
prolonged periods (8 months) if kept away from direct sunlight.
IV,48. POTA"SSIUM
Discussion. Potassium may be determined in one of the following
forms:
A. Potassium chloroplatinate; K2[PtffiaJ. This method is applic-
able only to those potassium compounds which can be completely
converted into potassium chloride by evaporation with hydrochloric '
acid (as by technique of Sect,ion IV, 47A), bec~use it is only from a
solution containing chloride that potassium can be completely
precipitated as K 2[PtCI 6] by chloroplatinic acid solution. Am-
monium sa.lts and allll}etais other than sodium and potassium must
be removed, as must also sulphate, phosphate and similar radicals.
Sodium chloroplatinate is soluble in 80 per ceJ?-t alcohol, hence this
method provides a means of separation of sodium from potassium.
lpdeed, potassium was frequently determine.d by this met'hod when
a separation from sodium was required. In recent years- tbis
separation has been. superseded by the procedure employing the less
expensive perchloriG acid (see B below).
The potassium chloroplatinate may be reduced by magnesium to
the metal and weighed as such. If this method is adopted, potas-
sium may be estimated in the presence of phosphates,. sulphates,
calcium, magnesium and many oth~r metals.
B. Potassium perchlorate, KCl0 4• This method employs the
comparatively inexpensive perchloric acid as the precipitating
reagent. All metals and ammonium, other than those of the alkali
metals, should be absent; the metals should be in the form of
chlorides. Under these. ccmditions the procedure may be employed
for the quantitative separation of potassium, sodium and lithium.
The three metals, as chlorides, are converted into the perchlorates
by evaporation with perchloric acid (DANGER, see Section IV. 70).
The mixed perchlorates are treated with a mixture of equal volumes
of anhydrous 'n-butyl alcohol (b.p. 116° to 118°C/76(1 mm.) and.
absolute ethyl acetate. This dissolves the sodium and lithium
perchlorates leaving the potassium perchlorate, which can be
weighed. The ethyl acetate in the washings is removed by evapora-
tion, and the sodium is precipitated as sodium chloride by the
addition of a solution of hydrogen chillride in n-butyl alcohol.
This is filtered off and weighed. Tl;e filtrate from the sodium is
evaporated, and the lithium is determined as sulphate.
If potassium alone is to be determined, many other metals may be
present as nitrates or chlorides.; these include those of the alkaline
earth metals and magnesium, but ammonium salts and sulphates
must be absent. In this instance it is ·sufficient to use 98 per cent
alcohol saturated with potassium perchlorate and containing a little
perchloric acid to remove the perchlorates of the metals other than
sodium. This method is also fairly accurate when applied to a
mixture of sodium and potassium chlorides.
, C. Potassium sulphate, K2S04• The experimental details are
similar to those given in Section IV, 47A. The method is applicable
to any potassium compound of any volatile acid. All other metals
must be absent. The temperature of ignition must not exceed
800 e owing to the danger of loss by v.olatilisation above this
0
temperature. .
D. Dipotassium sodium cobaltinitrite, K2Na[Co(N0 216J,H2o. By
precipitation of potassium solutions with sodium cobaltinitrite
reagent under the experimental conditions given below, a quantita-
tive yield of dipotassium sodium cobaltinitrite is obtained. This
method, unlike the perchlorate procedure, is applicable in the
presence of sulphate.
•
A. Determination of potassium as chloroplatinate. Pro-
cedure. Weigh out accurately about 0·25 g. of the mixed
sodium and potassium chlorides (1) into a small porcelain dish,
and dissolve inn 5 to 10 ml. of water. Treat with sufficient
chloroplatinic acid reagent (2) to react completely with both
the sodium and potassium. The necessary amount to u~e is
readily calculated by assuming that the mixed chlorides are
wholly sodium chloride-l·7 ml. of the reagent will precipitate
0·1 g. of sodium chloride. The dilution of the resultant
solution should be such that wh.en heated on the water bath,
any precipitate that may have formed dissolves completely;
this is to prevent occlusion of mother liquor in a mass of
crystals suddenly formed. Evaporate nearly to dryness on
the water bath, i.e., until the solution is syrupy but solidifies
on cooling. (Complete evaporation to dryness will,dehydrate
the sodium salt and render it less soluble in alcohol.) Add 10
ml. of 80 per cent (l::iy volume) ethyl alcohol (3) to the cold
dish, and break up the mass of crystals into a fine powder by
means of a stirring rod having a flattened end. Decant the
supernatant liquid through a weighed sintered glass or porous
porcelain crucible (4) . Repea t. this trea tmen t with 2 ml.
portions of 80 per cent alcohol, grinding up the precipitate
thoroughly each time, until the alcohol runs through entirely
colourle?s and the residue is golden-yellow in colour; .an
orange-red colour indicates incomplete removal of the sodium
salt. Transfer the residue to the filter crucible with the aid
of a little 80 per cent alcohol, and drain off most of the
572 QlIantitatit'e Inorganic Analysis
alcohol. Dry at 135°C to. constant weigh't: the crucible
should be covered to avoid loss by dectepitat.ion. Weigh as
K 2 [PtCl sJ·
Notes. 1. For practice in this procedure, the student may employ either
A.R. potassium chloride or an artificial mixture of, say, equal weights of A.R.
sodium and potassium chlorides.
2. The chloroplatinic acid reagent is prepared hy dissolving I g. of chloro-
platinic acid in lU ml. of water.
3. This may. be prepared by' mixing 5 volumes of 95 per cent alcohol with 1
volume of water.
4. AJI washings must be kept, and the platinum contaioed in them subse-
quently recovered. .
Weighing of potassium chloroplatinate as metallic platinum.
As a check on the result obtained by the aboye method .. the
amount of platinum contained in the precipitate may be
determined. Dissolve the precipitate by pouring hot water
over it, and transfer the filtrate and washings quantitatively
to a small beaker. Add 2 ml. of concentrated hydrochloric
acid, followed by approximately 0·5 g. of magnesium ribbon
(previously washed with water). Stir the solution, and hold
the ribbon at the bottom of the beaker by means of a glass
rod with a ft.attened end. When the magnesium has nearly
disappeared, add a few ml. of dilute hydrochloric acid, and
allow the plq_tinum to settle. If the reduction is complete,
the liquid is clear and colourless. To make sure add a little
more magnesium, and note whether the solution darkens.
Add dilute hydrochloric acid, boil to dissolve any basic salts
which lllay b~ present, collect the platinum on a small filter
paper, wash with water until fyee from chlorides, and ignite in a
a w~ighed porcelain or, preferably, platinum crucible -to
constant weight. Weigh the platinum, and calculate the
potassium equivalent by the proportion Pt === 2K.
B. Determination of potassium as perchlorate. Proced1ere.
Weigh out accurately about 0·3 g. of the mixed chlorides (1)
into a small Pyrex or platinum dish, and dissolve it in-5 to 10
m!. of water. Add 5 ml. of 20 per cent perchloric acid solu-
tion, evaporate to dryness on the steam bath, and then heat
carefully over a small ft.ame (or upon a low temperature hot
plate, ::} 350°C) until thick white fumes of perchloric acid
appear. Allow to cool, wash'down the walls of the dish with a
little water (5 to 10 mI.), add a further [) m!. of 20 per cent
pecchloric acid solution, and repeat the evaporation to ensure
complete conversion of the chlorides into the perchlorates.
Continue the s~cond evaporation until the contents of the
dish are almost dry, so as t'o eliminate all the perchloric acid.
Allow to cool. When cold, and not before (2), stir the con-
tents of the dish with 15' ml. of the special ~ash liquid (3).
Decant the supernatant liquid through a weighed filter
crucible (Gooch, sintered glass or porous porcelain). Repeat
. this with two 5 ml. portions of the wash liquid. Transfer the
precipitate quantitatively to the crucible with the aid 9f the
wash liquid. (About 50 m1. should suffice for the washing and
the transfer of the precipitate.) Dry at 130°C for 1 hour, and
weigh as KCI0 4 •
Notes. 1. For wacHee in this estimation, use either 0·3 g. of A.R. potassium
chloride or a mixture of. say. equal weights 0'1' A.R. potassium and sodium
chlorides.
2. On no account should the alcohol-perchlorate mixture be heated as a
dangerous explosion may result. For danger 3;ttending the use of pcrchloric
acid, see Section IV, 70.
3. The wash liquid consists of absolute alcohol containing 0·2 per cent of
perchloric acid and saturated with potassium perchlorate. This may be
prepared by adding 0·3 m~ of 60 per cent perchloric acid to 100 m!. of ab,:;olute
ethyl alcohol. shaking this with excess of finely-powdered potassium perchlor-
ate. and filtering through a dry tilter paper. •
B'. Note on the determination of potassium and sodium.
If the mixture supplied consists of the sodium and pota~sium
salts of volatile acids, it may be converted into the chlorides
by several evaporations with hydrochloric acid in a Main-
Smith crucible (compare Section IV, 4'iA). This will give the
totar chloride. The potassium is determined as the per-
chlorate as above, and therefrom the equivalent amount of
potassium chloride is computed. The sodium chloride is
obtained by difference.
D. Determination of potassium as dipotassium sodium
cobaltinitrite. Procedure. (Potassium in potassium sulphate.)
Weigh out accurately 0·1 g. of. say, A.R. potassium sulphate,
dissolve it in 10 to 15 m!. of water, add 2 drops of glacial
acetic acid, then excess (15 to 20 m1.) of the sodium cobaltini-
trite reagent (I). J!\raporate to dryness on the water bath.
Wash the residue by decantation with 5 per cent acetic acid,
tllen transfer it to a weighed sintered glass or porous porcelain
crucible, and wash it with cold water until free from acid. Dry
at 100°C to constant weight. Weigh as K 2 Na[Co(N0 2l 6J,H 20.
Calculate the percentage of potassium in potassium sulphate.
Note. 1. The sodium cobaltinitrite rea.gent is prepared as follows. Dissolve
22 g. of sodium nitrite in 40 m!. of water and add to it a solution q~ntaining
11·3 g. of cobalt acetate dissolved in 30 ml. of wa"ter and 10 ml. of glacial acetic
acid. Gently warm the mixture, remove the nitrous fumes which are evolved
by evacuation at the pump, and, aftoc standing for 24 hours, filter off the small
amount of yellow precipitate (due to the presence of potassium as impurity).
Imd make up the solution to 100 m!. Keep in a dark place.
IV, 49. LITHIUM
Discussion. Lithium, If present as a salt of a volatile acid, may
be determined as lithium sulphate, L480 4 by repeated evaporation
with sulphuric acid (see Section IV, 47A). No other elements may
be present. Here th.e change from the acid sulphate to the normal
sulphate takes place comparatively easily so that the addition of
ammonium carbonate is not essential.
When present with sodium and potassium as chlorides, the lithium
may be determined:
(a) by the n-butyl alcohol-ethyl acetate method (Section IV, 48B) ;
(b) by extraction of the 1ithium chloride with dry pyridine, amyl
alcohol;dioxan (diethylene oxide) or anhydrous acetone, sodium and
potassium chlorides being sparingly soluble iu. these solvents. The
best separation is achieved by the use of dioxan or anhydrous
acetone. The solubility of lithium chloride in dioxan is 2·8() g.
per litre at 100°C, an-d 38·6 g. per litre at 25°C in anhydrous acetone.
The splubility of sodium and potassium chlorides in these solvents is
negligible. The method utilising dioxan will be described.
Determination of litliium in the presence of sodium and
potassium. Procedure. Heat the mixed alkali chlorides at
120°C in a sintered glass crucible far 2 hours, and weigh after
cooling in a desiccator. Suspend the crucible by means of
platinum wires about 1 cm. above 20 to 30 ml. of pure dry
dioxan contained in a wide-mouthed 250 m!. flask, which is
attached to a reflux corrdenser having a calcium chloride guard
tube at the top. Reflu::c for at least 4 hours, remove the
crucible, heat at 120°C for 1 hour, cool, and weigh. Repeat
the process until the weight is constant. The loss in weight is
due to ,the lithium chloride.
IV, 50. AMMONIUM

Discussion. For the determination of ammopium by a gravimetric
procedllre, it must be present as the chloride ;' all other cations must
be absent. A little hydrochloric acid is added, followed by excess
of chloroplatinic acid reagent (see Section IV,.48A). The mixture
is evaporated almost to dryness on the water bath; the residue is
triturated with absolute' alcohol to remove excess of chloroplatinic
acid, and then transferred to a weighed filter crucible (Gooch,
sintered glass or porous porcelain). The crucible is dried at 130°C,
and th~ residual ammonium chloroplatinate. (NH4)2ptCls wt:,ighed.
(For details of experimental procedure, see Section IV, 48A.)
If the ammonium salt is present with other cations and
anions, a volumetric procedure' (see Chapter III) is usually
employed.
IV, 51. TUNGSTEN
Discussion. Tungsten, as tungstate, may be determined in one of
the following forms:
A. Tungsten trioxide. WO s• after precipitation with benzidine.
Benzidine forms a white sparingly soluble tungstate with solutions
. of tungstates. The precipitate is, however, filtered with difficulty,
and may tend to peptise during the washing process. If a little
sulphuric acid is present, the slightly soluble and crystalline benzi-
dine sulphate is also formed. This renders filtration facile. Upon
igniting the mixture of benzidine tungstate and .sulphate, only
tungstic oxide remains. , .
B. Tungsten trioxide. W0 3• after lIrecipitation with tannin and
antipyrine. Tungstic acid is incompletely precipitated from solu-
tions of tungstates by tannin. If, however, antipyrine is added to
the cold solution after treatment with excess of tannin, precipitation
is quantitative. This process effects a separation from aluminium,
and also from iron, chromium, manganese, zinc, cobalt and nickel
if a double precipitation is used.
For the use of cinchonine in the determination of tungsten, 'see
Section IV, 92,.1.
C. Tungsten" oxinate," W02(CuH60N)2' This method is applic-
able only when all other elements precipitable by oxiI].e in acetic
acid solution are absent.
A. Determination of tungsten as the trioxide (benzidine
method). Procedure. The solution of the tungstate (200
mI.) should contain about 0·3 g. of WO a, and be almost neutral
(1). Aad 10 ml: of O'IN sulphuric acid, and boil the solution.
Introduce immediately 30 ml. of the benzidine reagent (2),
and allow to cool to room temperature. Filter through a
weighed silica Gooch or porous porcelain crucible, and wash
with a solution prepared by diluting 10 ml. of the reagent to
300 ml. Dry the precipitate, and ignite. Weigh as W0 3 •
Notes. 1. For practice in this estimation, a solution prepared from A.R.
sodium tungstate may be employed.
2. The reagent is prepared by dissolving 2 g. of A.R. benzidine in 100 ml.
of 1 per cent hydrochloric acid.
B. ·Determination of tungsten as ·the trioxide (tannin-
antipyrine method). Procedure. The solution of tungstate
(200 to 25() ml.) should contain not more than 0·15 g. of WOs,
and be faintly ammoniacal. Add 6 to 7 ml. of concentrated
sulphuric acid and 7 to 8 g. of ammonium sulphate, and
heat to boiling. Treat with 6 ml. of 10 per cent tannin
solution, keep the mixture on the water bath for a few min-
utes, and allow to cool to room temperature. A flocculent
576 QuanUtative Inorganic Analysis
dark-brown precipitate separates. When cold, stir in 10 ml.
of a 10 per cent solution of antipyrine (" phenazone ").
Filter the precipitate through a weighed silica Gooch or
porous porcelain crucible (1), wash with the special wash
liquid (2), and ignite to constant weight. Weigh as W0 3 •
Notes. 1. The filtrate must be colourless. If it [s yellow, insufficient
antipyrine has been added.
2. The special wash liquid contains I mI. of concentrated sulphuric acid,
10 g. of ammonium sulphate, and 0·4 g.1>f antipyrine in 200 mI. of water.
C. Determination of tungsten with oxine. Procedure. The
sofution of tungstate (200.m!.) should contain about 0·05 g.
of W, and be slightly alkaline at neutral in reaction; if acid,
precipitation is subsequently effected with ammonium ace-
tate. Heat to. boiling, add a 4 per cent alcoholic solution of
oxine slowly and with constant stiFFing until present in slight
excess, and acidify with dilute acetic acid. • Filter through a
weighed filter crucible (Gooch, sinter~d glass or porous porce-
lain). Wash with hot water, and dry at 120°C to constant
weight. Weigh as W0 2 (C g H eONh
IV, 52. CHLORIDE
Discussion. This anion is determined as silver chloride, Agffi;
full details.a.re given in Section IV, c. Anions'which give silver salts
which are insoluble in dilute nitric acid must be absent; these
include bromide, iodide, thiocyanate, sulphide, thiosulphate, ferro-
cyanide and ferricyanide. Heavy metals interfere, and must be
removed by precipitation.
If the chloride is ins~luble, it is necessary to ~oil it with a large
excess of saturated sodium carbonate solution or, better, to fuse it
with sodium carbonate and extract the melt with water. In either
case the chloride passes into solution, and is estimated in tlte usual
way after acidification with nitric acid.
IV,53. BROMIDE
Discussion. This anion is determined as silver bromide, AgBr
by precipitation with silver nitrate solution in the presence of dilute
nitric acid. Silver bromide is less soluble in water than is the
chloride. The ..solubility of the former is 0·11 mg. per litre at 21°C as
compared wittI 1·54 mg. per litre for the latter, hence the p;-ocedure
for the determination of bromide is practically the same as that for
chloride. Protection from light is even more essential with the
bromide than with the chloride because of its greater sensitiveness.
IV,54. IODIDE
Discussion. This anion is usually determined by precipitation as
silver iodide, AgI. Silver iodide is the least soluble of the silver
.
halides; 1 litre of water dissolves 0·0035 mg. per litre at 21°C .
Coprecipitation and similar errors are more likely to occur with
iodide than with ~he other halides. Precipitation is therefore made
by adding a very dilute solution, say 0·05N, of silver nitrate slowly
and with constant stirring to .a dilute ammoniacal solution of the
iodide until precipitation is complete, and then adding nitric acid in
1 per cent by volume excess. The precipitate is collected in the
usual m~er, washed with 1 per cent nitric acid, arid finally with a
little water to remove nitric acid. Peptisation tends to occur with
excess of water. Other details of the estimation will be found in
Section IV, 5. •
Iodide Itlay also be determined by precipitation as paIladous
iodide, PdI2• Substances, such as alcohol, which cause reduction to
metallic palladium must be absent; bromides and chlorides are not
precipitated and therefore do not interfere. The precipitate is
insoluble in water and in dilute hydrochloric acid (1 : 99). The
reagent, palladous chloride, is expensive, and the method"'lS therefore
rarely employed except for gravimetric 'separations from other
halides.
Determination of iodide as palladous iodide. Procedure.
The iodide solution should contain 1 per cent by volume of
hydrochloric acid, and not more than 0·1 g. of iodide. Warm
to 70°C, and add palladous chloride solution, dropwise and
with stirring, until no more precipitate is formed. Allow the
solution to stand for 24 to 48 hours at 20° to 30°C, filter the
brownish-black precipitate of palladous iodide on a weighed
filter crucible (Gooch, sintered glass or porous porcelain), and
wash 4 times with warm water. Dry at 100°C for 1 hour,
and weigh as PdI 2 •
IV, 55. THIOCYANATE

This anion may be determined in one of the following forms:
A. Silver thiocyanate, AgCNS. This method is applicable only in
the absence of halides and cyanides. The dilute solution of alkali
thiocyanate is acidified with dilute nitric acid and treated in the
cold with a slight excess of 0 ·IN silver nitrate. After stirring well,
the precipitate is collected on a weighed filter crucible (Gooch,
sintered glass or porous porcelain), washed with water followed by a
little alcohol, and dried to constant weight at 130°C. The precipi-
_tate is weighed as AgCNS.
B. Cuprous thiocyanate, ~(CNS)2. The solution (100 m!.)
should be neutral or slightly acid (hydrochloric or sulphuric acid),
and contain not more than 0·1 g. eNS. It is saturated with sulphur
dioxide in the cold (or 50 ml. of freshly-prepared saturated sulphur-
ous acid ~olution added), and then treated dropwise and with con-
stant stirring with about 60 m!. of about 0'1N copper sulphate.
The mixture is. again saturated with sulphur dioxide (or 10 m!. of
V
saturated sulphurous acid solution added), allowed to stand for a
few hours, collected on a weighed filter crucible (Gooch, sintered
glass or porous porcelain), washed several times with cold water
containing sulphurous acid until the copper is removed (potassium
ferrocyanide test), and finally once with alcohol. The precipitate is
dried at 110° to 120°C to constant weight, and weighed as Cu 2(CNS)2'
C. Barium sulphate, BaS04 • The thiocyanate is oxidised with
bromine water to sulphate, and the latter is determined by precipi-
tation as barium sulphate. All other compounds cont.fJ.ining sulphur
must be absent. The alkali thiocyanate solution- is treated with
excess of bromine water, heated for 1 hour on the water bath, the
solution acidified with hydrochloric acid, and the sulphuric acid
precipitated and weighed as BaS04 (see Section lV, 6),
IV, 56. CYANIDE

Discussion. This anion -is determined as silver cyanide, AgCN ;
the experimental details are similar to those given for Chloride except
that, owing to the volatility of hydrocyanic acid, the solution must
not be heated. The cold solution of alkali cyanide is treated with an
excess of silver nitrate solution, faintly acidified with nitric acid, the
precipitate allowed to settle"collected on a weighed filter crucible,
and weighed as AgCN after drying at 1l0°C.
IV, 57. FLUORIDE

Discussion. This anion may be determined in one of the following
forms:
A. Lead chlorofluoride, PbCIF. The details of this procedure are
given in Section m. 39. The precipitate is, however, filtered
t,hrough a weighed filter crucible (Gooch, sintered glass or porous
porcelain), and weighed as PbCIF after drying at 130° to 140°C.
B. Calcium fluoride. CaF2. Fluorides, in the absence of phos-
phates, silicates and other radicals that give insoluble calcium salts,
may be determined as calcium fluoride by precipitation with calcium
chloride solution. The precipitate produced in acid or neutral
solution is gelatinous and difficult to filter. For this reason a little
sodium carbonate solution is often added before precipitation; the
precipitate of calcium fluoride and calcium carbonate is more readily
filtered. The calcium carbonate is subsequently removed by
extraction with dilute acetic acid. A simpler procedure is to carry
out the precipitation in ammoniacal solution; the resultant calcium
fluoride is much denser, and comparatively easily filtered and
washed.
C. Triphenyltin fluoride, (C6HS) aSnF, Triphenyltin chloride
reQgent precipitates fluorides quantitatively as the corresponding
fluoride. The precipitate is crystalline, easily filtered and quickly
washed, and is quite stable. Owing to the insolubility of the
reagent in water, precipitation is carried out in 60 to 70 per cent
alcoholic solution and washing is effected with an alcoholic solution
of the reagent saturated with triphenyltin fluoride.
The conversion factor is particularly small (0'05149), so that the
method is well adapted for the estimation of small quantities of
fluorides. The maximum amount that can be conveniently handled
is 0·04 g. of F. The solution sho.uld have a pH of 5 to 7; if acid
.some fluorine will be lost on heating to boiling, and if basic triphenyl-
tin hydroxide will be precipitated along with the fluoride. Metals
other than those of the alkalis should preferably be absent; the
alkali metals may be removed by washing the precipitate several
times with the alcohol wash solution, followed by cold water. Small
quantities of nitrates, chlorides, bromides, iodides and sulphates do
not interfere, but silicates (with ammoniacal zinc hydroxide),
phosphates (with silver nitrate) and carbonates must be removed
before precipitation. Carbonate is best removed by neutralising
with dilute nitric acid (to phenolphthalein), and boiling off the carbon
dioxide. A disadvantage of the' method is that the reagent is
expensive.
B: Determination of fluoride as calcium fluoride.
Procedure. The alkali fluoride solution should have a volume
of 75 to 100 ml., and contain up to 0'3 g. of F. Add 7 ml. of
concentrated ammonia solution and 20 ml. of 0'5M calcium
chloride in the cold, slowly,and with stirring. Boil the solu-
tion for 1 to 2 minutes, allow the precipitate to settle, and
decant the mother liquor through a quantitative filter paper.
Add 20 m!. of water. and 1 m!. of concentrated ammonia
solution, boil, and decant as before: repeat the process.
Transfer the precipitate to the filter, and wash 3 to 4 times with
hot water. Place the filter and precipitate in a weighed
Main-Smith crucible, bum off the filter paper, and ignite at
500 0 to 600°C (dull red heat) to constant weight. Weigh as
CaF 3' As an additional check, convert the calcium fluoride
into calcium sulphate by two or three evaporations with con-
centrated sulphuric acid, ignite at about 500°C and weigh as
CaS0 4 ; if this is carried out, the initial ignition must be
carried out in a platinum crucible.
C. Determination of fluoride as triphenyltin fluoride.
Procedure. The solution (say, 25 ml.) should contain not more
than 0·04 g. of F, and be almost neutral (1). Add 95 per cent
ethyl alcohol to the aqueous solution of the fluoride so that it
comprises about 60 to 70 per cent o~ the final volume. Heat
to boiling and treat with twice the calculated quantity of the
reagent (2) diluted with an equal volume of 95 per cent
alcohol and also heated to boiling. The latter is run slowly
into the hot fluoride solution with vigorous stirring, and the
whole again heated to the boiling point. Remove the source
of heat, and continue the stirring until the solution has cooled
somewhat (3). Allow to stand overnight, and cool for 1 hour
in ice (4). Filter through a weighed sin~ered glass or porous
porcelain crucible, wash with 95' per cent ethyl alcohol which
has been saturated with triphenyltin fluoride (about 50 mI.).
Dry for 30 minutes at 110°C, cool in a desiccator, and weigh as
(C6H5) aSnF .
Notes. 1. For practice in this estimation, A.R. sodium fluoride (obtainable
from the Mallinckrodt Chemical Works, U.S.A.; see, however, Appendix,
Section A, 9) may be used. If desired, a sample of pure sodium fluoride may be
prepared as follows. Treat A.R. anhydrous sodium carbonate with an excess
of A.R. hydrofluoric acid in a platinum dish, and allow to stand for a few
hours. Remove the excess of acid by heating (fume cupboard), allow to cool,
and add more acid. Mix thoroughly with a platinum spatula, heat the dish
gently at first, and then strongly until the sodium fluoride is entirely fused.
Pulverise in an agate mortar, dry the powder in platinum at 1l0°C, and store
in a desiccator over calcium chloride.
2. The reagent is prepared by shaking vigorously 4·0 g. of triphenyltin
chloride (obtainable from the Eastman Kodak Co.) with ,200 m!. of 95 per
cent ethyl alcohol; filter from the small undissolved residue. This is prac-
tically a saturated solution.
About 55 mi. of this reagent are required for 0·04 g. F. ,
3. If the quantity of fluoride is large, precipitation commences in about
a minute after the addition of the reagent as white crystals, but with small
quantities it does not take place until the solution has cooled to room
temperature.
4. This is unnecessary if the amount of fluoride is large and the total volume
of the solution small.
IV,58. CHLORATE
Discussion. The chlorate is reduced to-chloride, and the latter is
determined as silver chloride, AgCl. T?e reduction may be per-
formed with ferrous sulphate solution, sulphur dioxide or zinc.
Alkali chlorates may be quantitatively converted into chlorides by
three evaporations with concentrated hydrochloric acid, or by
evaporation with three times the weight of ammonium chloride;
the operations are conveniently carried out in a Main-Smith crucible·
(Section II, 36B).
Determination of chlorate as silver chloride. Procedure.
The chlorate solution should have a volume of about 100 ml.,
and contain about 0·2 g. of CI0 3 • Add. 50 ml. of a 10 per cent
soilltion of A.R. crystallised ferrous sulphate, heat with
constant stirring to the boiling point, and maintain the boiling
for 15 minutes. Allow to cool, add nitric acid until the
precipitated basic ferric salt is dissolved, precipitate the
chloride by means of silver nitrate solution, and collect an.d
weigh as AgCI after the usual treatment (Section IV, 5).
Alternatively, treat the chlorate solution with excess of
sulphur dioxide, boil the solution to remove the excess of the
gas, render slightly acid with nitric acid, and precipitate the
silver chloride as above.
For the reduction with zinc, render the chlorate solution
strongly acid with acetic acid, add excess of zinc, and boil the
mixture for 1 hour. Dissolve the excess of unused zinc with
nitric aCid, filter, and treat the filtrate with silver nitrate in
the usual manner.
Note. Hypochlorites and chlorites may be reduced to chlorides
with sulphur dioxide, and determined in the same way.
IV, 59. PERCHLORATE

Discussion. Perchlorates are not reduced by ferrous sulphate
solution, sulphurous acid or by repeated evaporation with concen-
trated hydrochloric acid; reduction occurs, however, with titanous
sulphate solution. Ignition of p')rchlorates with ammonium chloride
in a platinum crucible or in a porcelain crucible in the presence of a
little platinum powder results in reduction to the chlorides (the
platinum acts as a catalyst), which may be determined in the usual
manner. Losses occur when perchlorates are ignited alone.
I Perchlorates may also be determined as potassium perchlorate,
KCl04• Sodiunlperchlorate is estimated directly by precipitation
with concentrated potassium acetate solution. The sparingly
soluble perchlorates of the heavy metals are first precipitated with
excess of sodium ~arbonate solution, filtered, and the filtrate (con-
taining sodium perchlorate) evaporated to dryness on the water
bath. The cold residue is extracted six times with 5 ml. portions
of absolute alcohol, and the filtrate carefully evaporated on the water
bath. It is essential that the alcoholic solution be neutral or alka-
line in reaction, otherwise a dangerous explosion may result upon
evaporation. The residue is dissolved in 20 ml. of water, and
precipitated as KCI0 4 with potassium acetate solution.
A. Determination of perchlorate as silver chloride.
Procedure. The perchlorate. if in the form of a solution, is
evaporated to dryness on the water bath; otherwise the solid
perchlorate is used directly. Intimately mix about 0-4 g.
of the pe~chlorate (1) with 1-5 g. of A.R. ammonium chloride
in a platinum crucible covered with a watch glass or lid, ignite
gently until fuming ceases and continue the heating for 1
hour. Do not fuse the resultant chloride, as the crucible
'may be attacked. Repeat the ignition with another 1·5 g.
of ammonium chloride. Dissolve the residue in.a little water,
filter through a small quantitative filter paper to remove any
platinum powder which may be present, and estimate the
chloride in the filtrate as silver chloride (Section IV, 5).
Note. 1. For practice in this estimation, the student may employ A.R.
potassium perchlorate.
B. Determination of perchlorate as potassium perchlorate.

Procedure. The perchlorate solution should be in the form of
the sodium salt (see Discussion above}, have a volume of
about 25 mI., and contain about 0·4 g. of C10 4 • Warm the
solution to 80° to 90°C, and treat with a slight excess of a ,cold
saturated solution of potassium acetate. Allow the solution
to cool, and after 1 hour collect the precipitate on a weighed
filter crucible (Gooch, sintered glass or porous porcelain).
Wash twice with 0'05M potassium acetate, then 4 times with
absolute alcohol (1), and draw air through the precipitate
until it appears dry. Dry at llO°C to constant weight, and
weigh as KCI0 4 •
Note. 1. It is preferable to use the special wash liquid described in Section
IV, 48B, Note :l.
IV,60. IODATE
DiSC1tSsion. Iodates are readily reduced by sulphurous acid to
iodides; the latter are determined by precipitation with silver
nitrate solution as silver iodide, Agl. Iodates cannot be converted
quantitatively into iodides by ignition, for the decomposition takes
place at a temperature at which the iodide is appreciably volatile.
Periodates are also reduced by sulphurous acid, and may therefore
be similarly determined. Similar remarks apply to bromates; these
are ultimately weighed as silver bromide, AgBr.
Detennination of iodate as silver iodide. Procedure. Acidify
the iodate solution (100 ml. containing ca. 0·3 g. of lOa) (1)
with sulphuric acid, and pass in sulphur dioxide (or add a
freshly-prepared sat:urated solution of sulphurous acid) until
the solution which at first becomes yellow, on account of the
separation of iodine, is again colourless. Boil off the excess.
of sulphur dioxide, and precipitate the iodide with dilute silver
nitrate solution as described in Section IV, 54. Weigh as
AgI.
Note. 1. For practice in this determination, the student may ~mploy A.R.
putassium iouate.
IV, 61. SULPHATE
Discussion. Full details for the determination of soluble sul-
phatl's as barium sulphate, BaS04 are given in Section IV. 6. For
sulphates which are insoluble in water and acids, it is best to mix
the finely-powderNi solid with 6 to 12 times its bulk of anhydrous
sodium carbonate in a platinum crucible set in an asbestos shield
(Section II, 38), heat the covered crucible slowly to fusion, ami main-
tain in the fused state for 15 minutes. The melt is extracted with
water, the solution filtered, the residue washed with hot 1 per cent
sodium carbonate solution, and the cold filtrate carefully acidified
with hydrochloric acid (to methyl orange). The sulphate is deter-
mined as in Section IV, 6. If silica is present, the acid filtrate is
evaporated to dryness on the water bath to dehydrate the silica;
a little concentrated hydrochloric acid is added, followed by hot
water. The silica is filtered off, and the sulphate determined in the
filtrate-.
A comparatively rapid method for the determination of sulphate,
which is particularly suitable for routine estimations, is
based upon the precipitation as BaSO ol in the presence of
picric acid (C. H. LindIsIy, 1936). Prolonged digestion
(compare Section IV, 6) is unnecessary, and the solution may
be filtered after about 10 mihutes. The detailed procedure is
as follows. The slightly acidified sulphate solution is treated
with 1 tq 5 m!. of saturated picric acid solution according to
the sulphate content (I), a slight excess of barium chloride
solution is added rapidly (2), and the solution gently boiled
for 5 to 10 minutes. The precipitate may then be filtered
off, thoroughly washed, and weighed as in Section lV, 6.
Notes. 1. If large' amounts of sodium salts are present, 10 to 25 ml. of the
saturated picric acid solution are necessary.
2. In this method there is no advantage in adding the barium chloride
solution slowly.
IV,62. SULPHIDE
Discussion. The determination of sulphur in insoluble sulphides
by wet and dry processes and subsequenflveighing as barium sul-
phate, BaS04 , is fully dealt with in Section IV, 7. Other methods.
which are particularly applicable to soluble sulphides, include
oxidation in alkaline solution with sodium hypochlorite or hydrogen.
peroxide. '" followed by the estimation of sulphate as barium sulphate.
IV,63. SULpmTE
Discussion. Sulphites may be readily converted into sulphates by
boiling with excess of bromine water, sodium hypochlorite, sodium
hypobromite or ammoniacal hydrogen peroxide (equal volumes of
20-volume hydrogen peroxide and 1 : 1 ammonia solution). The
excess of the reagent is decomposed by boiling, the solution acidified
with hydrochloric acid, precipitated with barium chloride solution,
and the barium sulphate collected and weighed in the usual manner
(Section IV, 6).
*Ammoniacal hvdrogen peroxi,1e is lI~nally emflloyed. The hy,lro,(en
:peroxide must be free hom sulphate.
584 Quantitative Inorgam'c Analysis
IV,64. THIOSULPHATE
Discussion. Thiosulphates are oxidised to sulphate by methods
similar to those described for sulphitcs (Section IV, 63). The sul-
phate is then determined as barium sulphate. One molecule of
thiosulphate corresponds to two molecules of barium sulphate.
IV,65. PHOSPHATE
Discussion. Orthophosphates may be precipitated as magnesium
ammonium phosphate, MgNH4P0 4,6H20, by magnesium chloride and
ammonium chloride in ammoniacal solution (" magnesia" reagent).
Most elements, other than those of the alkalis, interfere, however, by
giving precipitates with "magnesia mixture." It is therefore
necessary in the majority of cases to separate the phosphate first
from interfering substances. This may be readily effected by preci-
pitation as ammonium phosphomolybdate. (NH4)aP04,12MoOa>
2HN0 3,H20, with excess of ammonium molybdate in warm nitric
acid solution: arsenic, vanadium, titanium, zirconium, silica, and
excessive amounts of ammonium salts interfere. The precipitate
thus obtained is dissolved in dilute ammonia,solution, and the phos-
phate is then precipitated as magnesium ammonium phosphate. A
double precipitation of the latter is necessary in order to obtain a
precipitate entirely free from molybdate.
If interfering elements are absent, the original precipi~ate of
ammonium phosphomolybdate may be weighed as phosphomolybdic
anhydride, P20s,24Mo03, after gentle ignition. For practice in this
simple estimation, the student may determine the percentage of
phosphate in anhydrous A.R. sodium phosphafe, Na 2HP0 4• (Most
of the experimental details are given in Section III, 146.) The
precipitate of ammonium phosphomolybdate should, however, be
filtered through a weighed Gooch or porous porcelain crucible, and
washed with the special wash liquid (1) until only a faint brown
colour is given when 5 ml. of the washings are tested with potassium
ferrocyanide solution. The precipitate is dried at 280 to 300°C 0
(electric oven) to constant weight. Alternatively, the precipitate is

aried in the steam oven, then placed in. a larger nickel crucible
resting upon a pad of ignited asbestos, 2 mm. thick, covering the
bottom of the crucible; the asbestos may be replaced by a porcelain
plate of a Gooch crucible, The filter crucible is covered with a watch
glass, and the nickel crucible gradually heated until the bottom is at
a dull red heat. In either case, when the precipitate has acquired a
homogeneous bluish-black colour, it is allowed to copl in a desiccator
(it is appreciably hygroscopic), and weighed as Pps,24Mo0 3 •
~ote. 1. Prepared by dissolving 10 g. of ammonium nitrate in water, 'ldding
8 mL of concentrated nitric acid, and diluting to 200 roL
A. Determination of phosphate as magnesium pyrophos-
phate. Procedure. To a neutral or weakly acid solution (50
to 100 ml.) of the phosphate, containing not more than 0·1 g.
of P205 and free from interfering elements (1), add 3 m!. of
concentrated hydrochloric acid and a few drops of methyl
red indicator. Introduce 25 mt. of magnesia mixture
(Section IV, 20B, Note 2), followed by pure concentrated am-
monia solution slowly whilst stirring the solution vigorously
until the indicator turns yellow. The procedure from this
stage is the same as described for the determination of Mag-
nesium in Section IV, 11, except that when carrying out the
reprecipitation from the hydrochloric acid solution 2 ml. gf
'the magnesia mixture are added instead of the 1 m!. of
ammonium phosphate solution. Weigh as Mg 2P t 0 7 •
about 0·4 g., accurately weighed, of A.R. anhydrous Na.HPO, ill 100 tnl. of
water.
B: Determination of phosphate as magnesium pyrophos..
phate after' a preliminary separation as ammonium phospho-
molybdate. Procedure. For practice in this estimation, it is
recommended that basic sla:g (I) be employed. Weigh out
accurately about 2·5 g. of the finely-powdered basic slag into a
10 cm. porcela,in or Pyrex basin. Add 20 m!. of concentrateq
sulphuric acid, and mix to a smooth paste. Heat the mixture
on a sand bath (or hot plate) in the fume cupboard until
white fumes begin to appear, and then keep it in this condition
for 45 minutes with stirring at intervals. Avoid strong
fuming of the sulphuric acid. Allow to cool, dilute carefully
with water, and, when cold, filter into a 250 ml. volumetric
flask, wash the residue tho toughly with cold water, and make
up to the mark. Measure out 50 ml. of the solution (do not
allow the solut·ion to stand too long as calcium sulphate is
liable to crystallise out) into a 400 ml. beaker, add 10 ml. of
concentrated nitric acid, followed by 1 : I-ammonia solution
until a faint permanent precipjtate appears. Redissolve the
precipitate by the addition of a few drops of concentrated
nitric acid, heat to 40° to 45°C ()1se thermometer), and treat
with 100 ml. of the ammonium molybdate reagent (Section
nI, 146) previously heated to the same temperature, slowly
and with stirring; continue the stirring for about 10 minutes.
Allow the precipitate to stand for a few hours, but preferably
overnight. Filter through a Whatman No. 40 or No. 540
filter paper, leaving as much as possible of the precipitate in
the beaker, an~ wash the precipitate moderately with 5 per
cent ammonium nitrate solution. Pour 10 ml of 1 : 2 am-
monia solution dropwise on to the paper to redissolve the
precipitate, wash with water, and collect the filtrate and
u*
washings in the original beaker which contains the bulk of
the precipitate. Render the ammoniacal solution slightly
acid (methyl red) with dilute hydrochloric acid, and introduce
25 m!. of magnesia mixture (Section IV, 20B, Note 2) dropwise
and stirring continuously. Next add 10 to 15 ml. of 1 : 1
ammonia solution slowly and with stirring. Allow the solu·
tion to stand for at least 4 hours, but preferably overnight.
Filter, without attempting to transfer the precipitate, wash
the vessel, residue and paper a few times with I : 20 ammonia
solution. Dissolve the precipitate on the filter in 25 m!. of
1 : I hydrochloric acid, and collect the solution in the original
beaker containing the bulk of the precipitate. \Vash the filter
paper thoroughly with warm 1 : 20 hydrochloric acid. Add to
the filtrate and washings, diluted to 100 mI., 2 m]. of magnesia
mixture; then reprecipitate the phosphate by the addition of
1 : 1 ammonia solution drop by drop and with thorough
stirring, and finally add 10 m). of the I : 1 ammonia solution
in excess. Allow to stand for 12 hours. Collect the precipi-
tate on a filter paper or in a weighed silica Gooch or porous
porcelain crucible, and wa'sh with cold 1 : 20 ammonia solu-
tion. Complete the determination as described in Section
IV. 11, and weigh as Mg2P20,.
Note. 1. Ri<'!sdale's" Basic Slag (Thomas Phosphate), No. 31 " (one of the
Analysed Samples for Students) or the Bureau of Analysed Samples" Basic
Slag' A ... (one of the British Chemical Standards) is suitable.
IV, 66. PHOSPWTE

DisCtfssioll. This anion may be determined in either of the
following forms:
A. Mercurous chloride. Hg2C4. The acid solution of phosphite
reduces mercuric chloride solution to calomel, which is weighed.
The reaction is :
2HgCI 2 + H aP0 3 + H 20 =' Hg 2CJ 2 + HaPO" + 2HCl,
whence Hg2CJ 2 =2 HaPO a.
B. Magnesium pyrophosphate, Mg2P207. The phosphite is oxi-
dised by ,nitric acid to phosphate, and the latter is determined as
magnesium pyrophosphate.
A. Determination of phosphite as mercurous chloride.
PrO(t,dHr.:. The phosphite solution (30 m!.) should contain
about 0·) g. (If H PU:1 Place 1)0 m!. of 3 per cent mercuric
chloride solution, 20 m!. of 10 per cent sodium acetate and 5
m!. of glacial acetic acid in a 250 ml. beaker, and add the
phosphite solutIOn dropwise, and with stirring, in the cold.
Allow to stand on a water bath at 30° to 45°C for 2 hours.
When cold, filter through a weighed filter crucible (Gooch,
sintered glass or porous porcelain), wash 2 or 3 times with 1
per cent hydrochloric acid, and then 4 times with warm water.
Dry at 100° to 105°e, and weigh as HgzCl z.
B. Determination of phosphite as magnesium pyrophos-
phate. Procedure. Treat the aqueous solution of the phos-
phite (100 ml.) with 5 ml. of concentrated nitric acid, evapor-
ate to a small volume on the water bath, add 1 ml. of fuming
nitric acid, and heat again. Dilute the solution, and precipi-
tate the phosphoric acid by magnesia mixture and ammonia
solution, and weigh as Mg zP z0 7 (Section IV, 65A).
IV, 67. HYPOPHOSPHITE

Discussion.' This anion is determined similarly to phosphite
either as mercurous chloride, HgzClz, or as magnesium pyrophos-
phate. MgJ» 207' In this case the reaction with mercuric chloride
solution is :
4HgCiz + H 3PO Z + 2H20 = 2Hg2Clz + H 3P0 4 -1- 4HCI,
so that 2HgzCI 2 == R aP0 2 •
IV,68. OXALATE
Discussion. The neutral solution of alkali oxalate is acidified
with acetic acid, heated to boiling, and precipitated with boiling
calcium chloride solution. After standing for 12 hours, the precipi-
tate is filtered off, washed with .hot water, and weighed either as
calcium carbonate, CaCO a• o"r as calcium oxide, CaO (for .details, see
Section IV. 10).
The following rapid method (J. Dick, 1929) yields results of
moderate accuracy. Precipitation of the oxalate is effected
in boiling solution containing a little ammonium chloride by
a hot solution of calcium chloride. The solution is allowed to
cool, treated with one-third of'its volume of rectified spirit,
and allowed to stand for 30 minutes. The precipitate is
washed by decantation through a weighed porous porcelain
or sintered glass crucible with warm water (50° to 60°C)
until the chloride reaction is negative. The calcium oxalate
is then transferred to the filter crucible, washed once with
cold water, 5 times with ·rectified spirit, and several times with
small volumes of anhydrous ether. The precipitate is
sucked dry at the pump for 10 minutes, the outside of the
crucible wipeCl dry with a clean linen cloth, and then left in a
vacuum desiccator for 10 minutes. ~t is weighed as
CaCp4,Hp.
IV,69. BORATE
piscussion. The most trustworthy methods for the determination
of borate are based upon its volatilisation as methyl borate, B (OCH 3) a;
this is achieved by acidifying the alkali or alkaline -earth borates and
heating with absolute methyl alcohol. (methanol). The methyl
borate is saponified and the boric acid fiXed in a weighed amount of
lime, which is again weighed. after ignition: Alternatively, the boric
acid is titrated after the addition of mannitol as described in Section
m, 17. For complete details, the reader is referred to other
treatises (1).
Note. 1. For example: W. F. Hillebrand and G. E. F. Lundell, Applied
Inorganic Analysis, 1929, p. 611; J. W. Mellor and H. W. Thompson, A
Treatise on Quantitative Inorganic Analysis, 1938, p. 659; w.n. Treadwell,
Tabellen und Vorschriften zur Quantitativen Analyse, 1938, p. 174; G. E. F.
Lundell, J. I. Hofmann and H. A. Bright, Chemical Analysis of Iron and Steel,
1931, p. 390.
IV,70. SILICATE
Discussion. Silicat~ may be divided into two classes: (a) those
(" soluble" silicates) which are decomposed by acids, such as hydro-
chloric acid, to form silicic acid and the salts (e.g., chlorides) of the
metals present, and (b) those (" insoluble" silicates) which are not
decomposed by any acid, except hydrofluoric acid. There are also
many silicates which are partially decomposed by acids; for our
purpose these will be included in class (b). Insoluble silicates are
generally fused with sodium carbonate, and the melt, which contains
the silicate in acid-decomposable form, is then treated with hydro-
chloric acid. The acid solution of the decomposed silicate is evapor-
ated to dryness_ on the water bath to separate the gelatinous silicic
acid Si02 ,xH 20 as insoluble silica Si0 2 ,yH 20 ; the residue is heated
at 110° to 120 C to partially dehydrate the silica and render it as
0
insoluble as possible. The residue is extracted with hot dilute

hydrochloric acid to remove salts of iron, aluminium, and other
metals which may be present. The greater portion of the silica
remains undissolved, and is filtered off. The filtrate is evaporated
to dryness, and the residue heated at llO° to 120°C as before in
order to render insoluble the small amount of silicic acid that has
escaped dehydration. The_ residue is treated with dilute hydro-
chloric acid as before, and the second portion of silic~ is filtered off
on a fresh filter. The two washed precipitates are combined, and
ignited in a platinum crucible at about lOOO°C to silicon dioxide,
Si02• and the latter is weighed. The ignited residue is not usually
pure silicon dioxide; it will generally contain small amounts of the
oxides of iron, aluminium, titanium, etc. The amount of impurity
may be determined, if desired, by treating the weighed residue in
the platinum crucible with an excess of hydrofluoric acid and a little
concentrated sulphuric acid. The silica is expelled as the volatile
"ilirnn tptr"flnnrine; the impurities (e.g., A120a and Fe20 a) are first
converted into the fluorides, which pass into the sulphates in contact
with the less volatile sulphuric acid, whilst the subsequent ignition
converts the sulphates back into oxides. Thus, for example:
Si02 + 6HF = H z[SiF6] + 2H zO; H z[SiF6] = SiF4 + 2HF;
AlzOs + 6HF = 2AIFs + 3H zO; 2AlFs + 3H zS04 = A1 2(SOJa +
6HF; Alz(S04h = AlzOs + 3S0 3•
The loss in weight therefore represents the amount of pure silicon
dioxide present.
Seventy per cent perchloric acid has been recommended for the
dehydration of silicic acid. The solution, containing hydrochloric
and nitric acid, is evaporated to strong fuming with excess of pel.'-
chioric acid, and heated for a further 15 to 20 minutes; water is
then added and the silica filtered off. A second evaporation is
necessary for the filtrate to separate the small amount of dissolved
silicic acid. The perchlorates of the metals, except that of potas-
sium, are readily soluble so that the method possesses certain advan-
tages. Organic compoUhds must, however, be absent, as dangerous
explosions may result.
EXPLOSIVE DANGERS OF PERCHLORIC ACID.-It must be
emphasised that aqueous and anhydrous perchloric acids are powerful
oxidising agents. •In the presence of carbon and easily oxidisable
organic compounds, a dangerous explosion may resu~t on heating with
the aqueous acid; in some cases explosion may result even in the cold
with the aqueous acid. Evaporations with perchloric acid should be
carried out in a fume cupboard, which is kept clean and free from
combustible materials. The acid ~hould be kept in a fire-proof store,
and reasonable care should be taken in handling it.
A. Detennination of silica in a "soluble" silicate.
Procedure. Weigh out accurately about 0·4 g. of the finely
powdered silicate (1) into a platinum or porcelain dish, add
10 to 15 ml. of water, and stir until the silicate is thoroughly
wet. Place the dish, covered with a clock glass or a Fisher
~t speedyvap" beaker cover, on the water bath, and add
gradually 25 ml. of 1 : 1 hydrochloric acid. The contents of
the dish must be continuously stirred with a glass rod; when
no gritty particles remain, the 'powder will have been com-
pletely decomposed. Evaporate the liquid to dryness: stir
the residue continuously and break up any lumps with the
glass rod. When the powder appears to be dry, place the
basin in an air oven at 1l0° to 120°C for 1 hour in order
to dehydrate the silica. Moisten the residue with 5 ml. of
concentrated hydrochloric acid, and bring the acid into
contact with the solid with the aid of a stirring rod. Add
75 m!. of water, rinse down the sides of the dish, and heat on a
steam bath for 10 to 20 minutes to assist in the solution of the
soluble salts. Filter off the separated silica on a Whatman
No. 41 or No. 541 filter paper. Wash the precipitate first
with warm 1 : 20 hydrochloric acid, and then with hot water
until free from chlorides. Pour the filtrate and washings into
the original dish, evaporate to dryness on the st~am bath,
and heat in an air oven at 1l0° to 120°C for 1 hour. Moisten
the residue with 5 ml. of concentrated bydrochloric acid, add
75 ml. of water, warm to extract soluble salts, and filter
through a fresh, but smaller, filter paper. Wash with warm
1 : 100 hydrochloric acid, and finally with a little hot water.
Fold up the' moist filters, and place them in a weighed
platinum crucible. Dry the paper with a small flame, char
the paper, and burn off the carbon over a low flame (Section
IT,36); take care that none of the fine powder is blown away.
When all the carbon has been oxidised, cover the crucible,
and heat for an hour at the full temperature of a Meker type
of burner in order to complete the dehydration. Allow to cool
in a desiccator, and weigh. Repeat the ignition, etc., until
the weight is constant.
To determine the exact Si0 2 content of the residue, moisten
it with 1 ml. of water, add 2 or 3 drops of concentrated sul-
phuric acid and about 5 m!. of the purest available (oA.R.)
hydrofluoric acid. Place the crucible in a radiator or air
bath (Section IT, 13F) and evaporate the hydrofluoric acid
in a fume cupboard with a small flame until the acid is com-
pletely expelled; the liquid should not be boiled. (The
crucible may also be directly heated, with a small luminous
flame.) Then increase the heat to volatilise the sulphuric
acid, and finally heat with a Mcker type burner for 15 minutes.
Allow to cool in a desiccator and weigh. Reheat to constant
weight. The loss in weight represents the weight of the
~ilic:a (2).
lVutes. 1. For practice in this estimation, powdered fused sodium silicate
may be used.
2. It is advisable to carry out a blank determination with the hydrofluoric
acid, and to allow for any non-volatile substances, if necessary.
B. Determination of silica in an I' insoluble" silicate.
Procedure. Weigh out accurately into a platinum crucible
about 1 g. of the finely-powdered dry silicate (1), add 6 times
the weight of anhydrous A.R. sodium carbonate (or, better,
of A.R. fusion mixture) and mix the solids thoroughly by
stirring with a thin rounded glass rod. Cover the mixture
with a little more of the carbonate, and_tben cover the
crucible. Heat the mixtun~ gradually until after about 20
minutes a tranquil melt is obtained; the cover of the crucible
is lifted occasionally to examine the contents. Maintain the
temperature of a quiet liquid fusion for about 30 JJ1inutes.
Allow to cool. Place the crucible and lid in a covered deep
porcelain or platinum basin (or in a large casserole), cover it
with water, and leave overnight, or warm on the water bath
until the contents are well disintegrated. When this is the
case, add very slowly about 20 m!. of concentrated hydro-
chloric acid by means of a pipette or bent funnel into the
covered vessel. Warm on the steam bath until the evolution
of carbon dioxide has ceased. Remove and rinse the cover
glass, crucible and lid, and evaporate the contents of the dish
to complete dryness on the steam bath, crushing all lumps with
a glass rod. Heat the residue for an hour at 110° to 120°C
to dehydrate the silica. Complete the estimation as described
in A.
Note. 1. Ridsdale's .. Felspar (Potash), No. 29" (one of the Analysed
Samples for Students) is suitable.
IV, 71. FLUOSILICATE

Discussion. The determination of this anion is of little practical
importance. The methods available for its estimation will, how-
ever, be outlined. Alkali fluosilicates ·are decomposed by heating
with sodium carbonate solution into a fluoride and silicic acid:
Na 2[SiF6] + 2Na CO + H 0 = 6NaF + H Si0 + 2C0
2 S 2 2 3 2•
Insoluble fluosilicates are brought into solution by fusion with 4

times the bulk of fu;;ion mixture, and extracting the melt with water.
In either case, the solution is treated with a considerable excess of
ammonium carbonate, warmed to 40°C:, and, after standing for 12
hours, the precipitated silicic acid is filtered off, and washed with 2
per cent ammonium carbonate solution. The filtrate contains a
little silicic acid, wllich may be removed by shaking with a little
freshly precipitated cadmium oxide. The fluoride in the filtrate is
determined as described in Section IV, 57.
If an acid solution of a fluosilicate is rendered faintly alkaline
with aqueous sodium hydroxide and then shaken with freshly pre-
cipitated cadmium oxide, all the silicic acid is adsorbed by the
suspension. The alkali fluoride is then estimated in the filtrate.
IV,72. FERROCYANIDE
No satisfactory gravimetric procedure is available. Volumetric
methods are described ill Chapter III.
IV,73. FERRICYANIDE
No satisfactory gravimetric method is availabie. For volumetric
methods, see Chapter III.
IV, 74. NITRITE

No satisfactory gravimetric procedure is available. Volumetric
methods are described in Chapter III.
IV, 'i5. NITRATE '

Discussion. The mono-acid base nitron.C 2oH 16N, forms a fairly
insoluble crystalline nitrate C2oH16N4,HNOa, (solubility is 0'099 g.
per litre at about 20°C) which can be used for the quantitative
estimation of nitrates (see Section I, 62, I). The sulphate and
acetate are soluble so that precipitation may be made in sulphuric
or acetic acid solution. Perchlorates (0'08 §.), iodides (0'17 g.,
thiocyanates (0'4 g.), chromates (0·6 g.), chlorates (1·2 g.) nitrites
(1·9 g.), bromides (6·!'g.), ferrocyanides, ferricyanides, oxalates,
and considelable quantities of chlorides interfere and shoul,d be
absent. The figures in parentheses are th~ approximate solubilities
of the nitron salts in grams per litre at about 20°C.
Determination of nitrate as nitron nitrate. Procedure.
The solution (75 to 100 ml.) should be neutral and contain
about 0·1 g. of NO a. Add 1 ml. of glacial acetic acid or 0·5 ml._
of 2N sulphuric acid, and heat the solution nearly to the
boiling point. Then introduce 10 to 12 ml. of the nitron
reagent (1), stir, and cool in ice-water for 2 hours. FiltE!r
through a weighed filter crucible (Gooch, sintered glass or
porous porcelain). Wash with 20 to 40 ml. of a cold saturated
solution of nitron nitrate"added in several portions, and drain
the precipitate well after each washing. Finally, wash twice
with 3 ml. portions of ice cold water. Dry at 105°C (1 hour
is usually required), and weigh as L2oH16N4,HNOa'
Note. 1. Prepare the reagent by dissolving 5 g. of nitron in 50 ml. of 5 per
cent acetic acid.
IV, 'i6. CARBONATE

Discussion. The carbonate is decomposed by dilute acid, and
either the loss in weight due to the escape 01 carbon dioxide deter-
mined (indirect method) or the carbon dioxide evolved is absorbed
in a suitable medium and the increase in weight of the absorbent
determined (direct method). The direct method gives more satis-
:factory results, and will therefore be described. The indirect method
is often employed, however, for samples containing relatively larg~
;lmolJnts of carbonate.
The decomposition of the carbona~e may be effected with dilute
hydrochloric acid; dilute perchloric acid or syrupy. phosphoric acid.
The last-named acid is perhaps the most convenient because of its
comparative non-volatility and the fact that the reaction can be
more easily controlled than with the other acids. If dilute hydro-
chloric acid is employed, a short water-cooled condenser should be
inserted between the decomposition flask and the absorption train
(see 'below).
Two absorbents are required, one for water vapour and the other
for. carbon dioxide. The absorbents for water vapour which are
generally employed are: (a) anhydrous calcium chloride (14 to 20
mesh), (b) magnesium perchlorate trihydrate (H dehydrite "), and
(c) anhydrous magnesium perchlorate (H anhydrone "). Both
(b) and (c) are preferable to (a); the former absorbs abo~t 35 per
cent and the latter about. 50 per cent of its weight of water. They
are, hQwever, expensive, and anhydrous calcium chloride is gener-
ally employed for ordinary work. The latter usually contains a
little free lime, which will absorb carbon dioxide <lIso; it is essential
to saturate the U-tube containing the calcium chloride with dry
carbon dioxide for several hours and then to displace .the carbon
dioxide by'a current of pure air.
The absorbents for carbon dioxide in general use are: (d) soda
lime (this is available also iIi the form of self-indicating granules,
" sofnolite " or as carbosorb " soda lime, which indicate when the
H
absorbent is exhausted), (e) soda lime-asbestos (the carbosorb "

H
variety gives a marked colour change and therefore indicates the

degree of exhaustion), and sodium hydroxide-asbestos (" ascarite ").
In all cases the carbon dioxide is absorbed in accordance with the
followi~g equation:
2NaOH + CO2 = Na2C0 3 + HzO.
Water is formed in the reaction, hence it is essential to fill one-
quarter or one-third of the tube
most removed from the carbon
dioxide train with a1l.Y of the des-
iccants referred to above (Fig. 4-2).
Determination of carbon dioxide

in a carbonate. Procedure. Fit up
the apparatus shown in Fig. 4-3.
A is a flask of about 100 ml.
capacity, B is a dropping funnel
containing 20 to 25 ml. of A.R. FIG. 4-2.
syrupy phosphoric acid, C is a soda-lime ,guard tube, D
is a bubbler containing syrupy phosphoric acid, E is a
U-tube containing calcium chloride which has been sat-
urated with catbon .dioxide and the residual carbon dioxide
displaced by air (this may be replaced by " dehydrite" or
" anhydrone," if available), F and G are soda lime or" sofno-
lite" tubes (these may be replaced by "carbosorb" or
Flq.4-3.
" ascarite "), and H is a guard U,tube containing the same

desiccant as in E. The U-tubes may be suspended by silver
wires attached to hooks on the glass or metal rod I, or by
some other means. All joints are made with short lengths of
stout-walled r!lbber tubing, and the two ends of the glass
tubing should be in contact, Rubber bungs are employed in
A, Band C. Before proceeding with the actual estimation,
make sure that the apparatus is gas tight.
Weigh out accurately 0,5 to 0·6 g, of the carbonate (1) into
the.flask A, which should be clean and dry. Remove the two
soda lime or " sofnolite .. tubes F and (1, wipe them with a
clean linen handkerchief or cloth, and leave them in the bal-
ance case 'for 45 minutes. Open the taps of the U-tubes
momentarily to the air in the balance case, and weigh them
separately. Replace them on the drying train; place 25 ml.
of A.R. syrupy phosphoric acid in B, and see that the appara-
tus is connected up as in Fig. 4-3. Open the taps of the U-
tubes. Run in sufficient phospho.ric acid from the tap funnel
to cover the solid in the flask (the 25 ml. will more than suffice).
Close the tap of the funnel and heat the flask carefully;
regulate the temperature so that not more'than 2 ~ubbles of
gas per second pass through the bubbler D. After about 30
to 40 minutes, the contents of the flask should be boiling;
boil for 2 to 3 minutes. Remove the flame, and immediately
attach a filter pump and a large bubbler (similar to D, and
containing syrupy phosphoric acid) to the end of the appara-
tus. Open the tap funnel, and draw air through the apparatus
at the rate of about 2 bubbles per second for 20 minutes.
Remove the tubes· F and G, close the taps, treat them as
before, and weigh them. From the increase in weight, calcu-
late the percentage .of CO 2 in the sample.
Note. 1. For practice in this estimation, the student may employ A.R.
calcium carbonate, or any of Ridsdale's " Limestone, Nos. 15, 15b, 15c, or
15d" (Analysed Samples (or Students).
ELECTROLYTIC DETERMINATIONS
IV, 77. Apparatus.-A. Electrolysis unit. The actual set-up
employed will vary from one laboratory' to another according
S
to the financial
M+___/ /--------------. resources and the
equipment otherwise
available. The circuit
will, however, be the
same in all cases, and
this will therefore be
considered first. A
circuit employing the
mains (200 to 240 volts
FIG. 4-4. or 110 volts D.C.) is
shown in Fig. 4-4; if
a D.C. supply is not to hand, some form of rectifier must
be employed for A.C. mains. M is the D.C. mains, Rl is a
large variable resistance (which may consist of a bank of
lamps), R2 is a smaller variable resistance, A is an ammeter
reading up to 10 amperes, V is a voltmeter reading up to 10 to
15 volts, E is the electrolysis vessel, and S is a switch. This
circuit is perfectly safe; there is no danger <_Jf the student
receiving any electrical shock in the usual operation at the
end of the electrolysis when the electrolyte is replaced by
water with the current still flowing. Many types of commer-
cial apparatus, incorporating this circuit, are available. * In
general, the two resistances, ammeter and voltmeter are
mounted on a board of insulating material, and two terminals
are provided for the electrolytic cell. The current may also
be supplied by a number of accumulators connected in series;
these should possess a sufficiently large ampere-hour capacity
* For example, Messrs. A. Gallenkamp and Co. Ltd., of London, England,
market two switch-boards for controlling current in electrolyses, one for use
with a 10 to 15 volt supply, and the other for operation with a 220 volt D.C.
supply mains. The Fisher Scientific Co, of Pennsylvania, U.S.A., supply an
.. electro-analyser OJ for use on an llO volt A.C. circuit; this apparatus is
complete with stirrer, stand for electrolysis cell, etc., and can be had in
multiple units, thus permitting up to 6 electrolytic determinations to be
carried out simultaneously. E. H. Sngent and Co., of Chicago, Ill., U.S.A.,
manufacture a similar apparatus for use on 110 or 220 volt mains. Atten-
tion is also directed to the very useful and comparatively inexpensive Slomin
.. electrolytic analyser" manufactured by E. H. Sargent and Co. of Chicago.
This small apparatus is entirely self-contained, portable, and requires only
connexion with the mains (IlO volts, 60 cycles A.C. or 220-240 volts 50 cycles
A.C.).
596
to supply a steady current during the whole period of the
electrolysis. In the author's laboratory six accumulators of
80 ampere-hours capacity are connected in series, and the
electrical connections are such as to permit any electrical
pressure up to 12 volts being obtained. These large accumu-
lators are for the low tension supply of the Chemistry Depart-
ment, and permit of several electro-analyses being conducted
simultaneously.
If the polarities of the terminals are not known, they may be
determined by touching the two wires from the terminals on
to paper moistened with potassium iodide solution; a brown
stain of iodine will form at the positive pole.
A very useful apparatus, and applicable with slight modifi-
cation to electrolytic, potentiometric and conductimetric
analyses, has been designed by H. ]. S. Sand* and is shown in
Fig. 4-5 ; the circuit is given in Fig. 4-6. The front panel of
FIG. 4-5.
• Manufactured by and obtainable from Griffin and Tatlock Ltd., Kemble
Street, London, W.C.2, England. The apparatus operates from a 12 or 6 volt
supply mains or battery. A car battery may be used; this can be charged
with the ordinary commercial A.C. mains charging set.
insulating material carries an ammeter A (0 to 10 amperes),
a voltmeter V (0 to 10 volts on either side of the central zero),
the main switch S for
- breaking the battery cir~
cuit, and a shunt SH which
",
£LECTROO~S
allows the rheostat R of 15
amperes carrying capacity
to be placed in shunt with
the battery :s,. for separa-
FIG. 4-6.
tion by graded or controlled
potential; the reversing
switch RS allows the outer electrode to be made either the
anode or cathode. At the top of the stand a motor is mounted,
which is controlled by a switch and the rheostat R 1• The driving
belt passes over- two pulleys to a hollow shaft to which a short
length of thick-walled rubber tubing I is attached. The tub-
ing is closed at the lower end by a stainless steel rod secured
to a V-grooved silver clutch, and is filled with mercury which
allows frictionless electrical contact with the inner electrode
to be maintained. The outer (stationaryT electrode is sup~
ported by a silver lined clamp 0, and the inner (rotating)
electrode -being provided with a silver or platinum rod at the
upper end, is merely pushed into the silver clutch by means of
which it is firmly held, for the rotation. The ring C is a beaker
support, and slides on two vertical steel rods, a quick release
for it being provided by the ring A, which fastens to a hook
on the stand. AE is an auxiliary electrode holder, for use in
supporting the auxiliary electrode required in refined analyses
by graded or controlled potential, electrolytic separations, etc.
B. Electrodes. These are generally made of platinum (or
platinum-iridium), although cathodes of tantalum are occa-
sionally employed for' the deposition of certai~ metals.
Silver electrodes (Fig. 4-10 a) are useful as cathodes for copper
and zinc determinations. Numerous types of electrodes have
been employed. The present tendency is to use gauze elec-
trodes, particularly for the cathode, because they are con-
sidered to be the most satisfactory. It is probable that the
interstices in the gauze bring about a much better circulation
of the electrolyte so far as conditions at the surface of the
electrode are concerned, the improved circulation serving to
reduce any local depletion of the electrolyte that might tend
to occur. It is probable that these gauze electrodes-may not
be in the possession of every laboratory, so that other, in some
cases, less expensive, electrodes will be described also.
The simplest apparatus is to employ a platinum dish as
cathode and a circular platinum plate as anode (Fig. 4-7 a) :
a circular platinum gauze +
electrode (Fig. 4-8 e) may
also be used. The filter
flask (Fig. 4-7 b) is for
the removal of the elec- \_ -'- I
trolyte after the electro- ..___..,
lysis. L. Classen (1927)
gives the following details:
diameter of dish 9 cm. ,; ( a) (b)
depth in the- centre 4·2 FIG. 4-7.
cm.; capacity about 250 ml.; weight 35 to 40 g.; the
inside surface should be matted (e.g., by sand blasting) so that
the deposit adheres more easily; surface area for 150 m!. is ca.
100 sq. cm., and for 180 ml. of liquid ca. 130 sq .. cm.; diameter
of anode 4'5 cm.; the electrodes should be used exclusively
for electrolytic determinations. For a dish 6·5 cm. in dia-
meter, the diameter of the anode should be 3·5 em.
Two other designs of cathode are shown in Fig. 4-8 a and
b; these are made of platinum sheet and wire. The anodes
c, d and e may be used with a, and the anode/with b. These
3"
7 "
II
6
/I
5
(0) (b) (0) (d) (e> (I>

[30 g.] [11 g.] [25 g.] [16 g.] [I6,g.] [16 g.]
FIG. 4·8.
are chiefly employed for stationary or slow electrolysis. The
cylinders may be provided with thin slits to facilitate the
circulation of the electrolyte. The" flag" type of electrodes
are shown in Fig. 4-9 a and b.
a"
6"
5"
/I
4
3"
2'
(b)
(a)
o [19g.j [9g.] (a) (b)

[16 g.] [28g.]
FIG. 4-9. FIG. 4-10. .
a" For routine work in rapid

electrolysis a platinum gauze
7" cylinder (Fig. 4-10 a) and the
associated gauze cylinder (Fig.
6 " 4-10 b) or spiral (Fig. 4-8 f) are
/I generally employed.
5
Sand's original type (1907) of
/I electrodes for rapid electrolysis
4
are shown diagrammatically in'
Fig. 4-11 a and b; a glass sleeve
3"
passes through b in order to
insulate the electrodes from one
2"
another. These are particularly
well adapted for very rapid
1" electrolysis.
(a) (b)
The electrodes shown in Figs.
0 [36 g.l [37 g.] 4-8, 4-9, 4-10 and 4-11 are
FIG. 4-11. drawn to one third the actual
size,. and the approximate weights are indicated. *
Sand's original electrodes, although yery convenient in use,
suffer from the disadvantage that they are expensive. A
7"
6 "
5"
4"
GLII.S5
3"
2"
1·"
0
(a) (6) (c)
lUg.] [12 g.]
FIG. 4-12.
modification, also due to Sand, which possesses most of the

advantages of the original design but requires comparatively
little platinum is shown in Fig. 4-12, together with the details
.
of construction.t The inner electrode is made of Pyrex glass,
• These are reproduced by the courtesy of Messrs. Johnson Matthey and
Co. Ltd., of Hatton Garden, London, E.C.I, England. They can of course be
modified to suit any particular purpose. Apparatus of smaller dimensions
down to .. micro" size can be obtained from this firm.
! The author is indebted to Dr. H. J. S. Sand for permitting him to repro-
duce details of the latest modification of these electrodes. The electrodes are
manufactured by Messrs. Griffin and Tatlock Ltd., Kemble Street, London,
W.C.2, England. They are drawn palf tpe actual size in tpe Figure.
and the platinum gauze and the four connecting wires are
lightly fused on to the. surface of the glass: a platinum wire
is wound round the glass to prevent "sticking" during
rotation.' The outer electrode (c in Fig. 4-12) is of platinum
gauze and weighs 9 to 12 g.; it is supported by means
of a Pyrex glass tripod, and a terminal is provided for fixing
into the silver-lined clamp 0 of the stand (Fig. 4-5). In use,
the stem of the inner electrode is slipped through the guide
tube of the tripod and fitted into the holder or clutch (Fig.
4-5 or Fig. 4-16); the outer electrode is pushed inside the
tripod and both it and the tripod are held by the cathode
clamp.
A mercury cathode is sometimes employed for special
operations. The most important of these is the separation
of aluminium from such metals as iron, manganese, chro-
mium, nickel, cobalt, zinc, molybdenum, cadmium, copper,
tin, bismuth and lead, which can, under suitable conditions,
be deposited on the mercury. The method is therefore of
particular value for the determination of aluminium in steels
and alloys. Three types of mercury cell are shown in Fig.
4-13. In (a) the platinum wire is sealed into the side of a lipped
(a) (b)
(0)
FIG. 4-13.
Pyrex beaker (250 ml.), whilst in (b) the'platinum wire is sealed

into the side tube, thus permitting the almost complete
separation of the aqueous and mercury layers. The platinum
terminal is, of course, covered with mercury. Apparatus (e)
(A. D. Melaven, 1930) is perhaps the most useful form; the
diagram is almost self-explanatory. The electrolysis vessel
contains the platinum anode (preferably of the rotating type)
immersed in the electrolyte. Electrical contact to the mer-
cury in the levelling bulb is made through a platinum wire
Gravimetric A nalysls 603
sealed into its side (alternatively, a pice of copper wire may dip
into the mercury); the mercury acts as cathode. When elec-
trolysis is complete the levelling bulb is lowered until the mer-
cury reaches the upper end of the stopcock bore; the stop-
cock is then turned through 180 degrees and the e~ectrolyte is
collected in a suitable vessel. It must be pointed out that in
addition to aluminium the following elements are' also not
removed by electrolysis in a mercury cathode cell: titaniuni,
vanadium, uranium, phosphorus, the alkaline earths and the
rare earths; the anolyte should be sulphuric acid of about
O'lN com;entration.
It is frequently necessary to calculate the current density:
For a smooth platinum surface, the estimation of the surface
area presents no particular difficulty. As a rough approxima-
tion the usual gauze electrode may be regarded as having an
effective area twice that of a plain foil electrode of the same
dimensions: this applies only if the surface of the foil
cathode next to the anode is effective in the deposition.
For a closer approximation of the effective area of a gauze
electrode, it is probably best to ·compute the surface area of
the wire used in making the gauze. The total length of wire
can be calculated from the number of meshes and the dimen-
sions of the electrode. The effective area will then be the
total length of the wire multiplied by 1Td, where d is the
diameter of the wire.
C. Electrode vessel. Tall-form beakers, without lip, are
generally employed to hold the solution to be electrolysed.
These should be of such size that there is· the smallest prac-
ticable volume of liquid between the cathode and the glass.
The apparatus should be so arranged that the beaker can be
easily removed without touching any other part
of the apparatus, and also so that"the electro-
lyte can be warmed, if necessary. To prevent
loss by spraying, split clock glasses with a
central hole (Fig. 4-14) are used for covering
the electrolytic vessel.
A form of electrode vessel, which .has been
employed for rapid electrolysis,
is shown in Fig. 4-15. At the end
GD FIG. 4-14.
of the electrolysis the solution may
be easily drawn off (and used in
other determinations, if de~ired)
and the electrodes washed without
interrupting the circuit. The FIG. 4-15.
disadvantages are: (i) the vessel cannot be heated easily,
and (ii) the liquid just above the tap does not readily mix
with the bulk of the solution.
D. Stirring. It has already been pointed out (Section I, 70)
that stirring of the electrolyte during electrolysis reduces
considerably the time required for deposition. J'he method
commonly employed is to rotate the inner electrode; this is
usually made the anode except in lead determinations (as
Pb0 2) when it is the cathode. Speeds up to 500 to 1000
revolutions per minute are permissible; those usually em-
ployed are 300 to 600 r.p.m. The rotation is generally ob-
tained by drive from an electric motor; a device is always
incorporated for varying the speed, for example, an electrical
resistance with a D.C. motor, pulleys, gears, etc. Some-
times an independent stirrer is employed, but this is not quite
so satisfactory as rotating one of the electrodes.
Many of the commercial apparatus (see Footnote in A) have
the stirring arrangement incorporated in the main stand. A
simple apparatus, utilising Sand's electrodes is shown in
Fig. 4-16. The inner electrode is attached through a silver
or silver-plated chuck and
rubber tubing to a pulley
shaft so that it can be
'2 I;J rotated; electrical con-
11 nexion is made through
.... mercury contained in the
rubber tube. The cathode
~ is separated from the
anode by a glass tube
surrounding the stem of'
the latter, and electrical
L'"\._
Circuit i s c 0 ;Ill pie ted
k:i . through the silver-plated

clamp holding the stem of
the cathode.
,j
)~
E. Use and care of elec-
trodes. Electrodes must be
free from· grease, other-
wise an adherent deposit
may not be obtained.
AO~ I:&! For this reason an elec-
trode should never be
FIG. 4·16. touched on the deposition
surface with the fingers; it sh.ould always be handled by the
platinum r.od .or wire attached t.o the main b.ody .of the
electr.ode. Platinum electr.odes are easily rendered grease
free by heating them t.o redness in a flame.
Bef.ore use electr.odes must be carefully cleaned .t.o rem.ove
apy previ.ous dep.opits. Dep.osits.of c.opper, silver, cadmium,
mercury, and many other metals can be removed by immersion
in dilute nitric acid (1 : 1), rinsing with water, then boiling
with fresh 1 : 1 nitric acid f.or 5. to 10 minutes, followed by a
final washing with \,Vater. C.opper may also be removed by
means .of a soluti.on c.omposed of 20 g. of trichl.oroacetic acid,
100 m!. of concentrated amm.onia soluti.on and 100 ml. .of
water. Dep.osits.of lead di.oxide are best removed by means
of 1 : 1 nitric acid containing a little hydrogen peroxide to
reduce the lead to the bivalent form; alcoh.ol or .oxalic acid
may replace the hydrogen peroxide. In all case,s it is. rec.om-
mended that the electrodes be heated to bright redness over
the colourless flame.of a Bunsen.or Mekertype burner bef.ore use,
When the electr.olysis is complete, the electrode with its
deposit is washed by a stream of water (ca. 100 ml.) from a
wash b.ottle .or .other supply while in place and with the current
still on; this is necessary since .otherwise the electr.olyte in
c.ontact with the electrode may dissolve some of the deposit.
The electrode is then disc.onnected and rinsed with alcohol*
.or, better, with pure (e.g., A.R.) acetone (15 t.o 20 mI.), and
then dried at 100° t.o 1l0°C f.or 5 to 10 minutes. In s.ome
instances the electrode may be dried by suspending it from a
hook clamped to the electrolysis apparatus, and waving 'a
flame bel.ow it. The drying should not be prolonged as there
may be some danger of .oxidising the metal dep.osit (e.g., with
c.opper). The electr.ode with its deposit is weighed after
co.oling t.o the lab.oratory temperature. In a few cases, there
is evidence that a small transfer .of ·platinum takes place. If
this is suspected, the electrode plus deposit should be weighed,
the deposit diss.olved, the electr.ode ·weighed again and the
difference taken. The .occurrence of an.ode c.orr.osi.on is m.ost
siinply detected .by weighing the anode before and after
electr.olysis .
• The use of ether after the alcohol washing is not recommended, since
ether frequently contains peroxides which may oxidise the metal. If suitable
precautions are taken to remove and to prevent the accumulation of peroxides,
e.g., by first shaking the ether with potassium permanganate solution or with
a weakly acid solution of ferrous sulphate followed by a little water, and then
storing in a dark place and including a copper spiral below the cork, ether may
be used.
606 Quantitative Inorganic. Analysis
The method adopted for the removal of the electrolyte
after the electrolysis is complete will naturally depend upon
the type of cell that is employed. With a platinum dish or
beaker, the electrolyte may be separated by means of the
siphon arrangement shown in Fig. 4-7 b, or by some equivalent
device. If the electrolysed solution is not to be preserved for
a subsequent determination, a very satisfactory method of
washing consists in siphoning the solution out of the electro-
lysis vessel while maintaining the liquid level in the vessel
by adding distilled water; the removal of the electrolyte is
shown by tlie ammeter needle dropping back towards zero.
If the beaker rests on a block of wood, the latter may be
displaced and the beaker lowered. In the commercial types
of apparatus referred to in A, a simple device for lowering the
beaker is usually incorporated. If the vessel is of the type
shown in Fig. 4-15, the liquid is simply run out by opening the
tap, and the electrode washed.
Electrolysis must never be carried out in acid chloride
solutions since in such solutions corrosion of the platinum
occurs, leading not only to erroneous results (the dissolved
platinum is plated out on the cathode) but also to injury of
the anode. * A number of metals (for example, zinc and bis-
muth) should not be deposited on a pure platinum surface.
These metals, particularly zinc,. appear to react with the
platinum in some way, for when they are dissolved off with
nitric acid the platinum surface is dulled or blackened.
Injury to the platinum can be prevented in this case by first
plating it with copper, and then depositing the metal <?n this
surface; alternatively, a silver gauze electrode (as Fig. 4-10 a)
may be used.
IV, 78. COPPER
Discussion. The general principles of electro-analysis have been
discussed in Sections I, 64 to I, 70; these should be mastered before
any determinations are made. Copper may be deposited from
either sulphuric or nitric acid solution, but, usually, a mixture of the
two acids is employed. If such a solution is electrolysed with.an
e.m.f. of 2 to 3 volts, the following reactions occur:
Cathode: Cu++ + 2E ~ Cu;
2H+ + 2E ~ H 2•
Anode: 20Ir -+ 1/20 2 + H 20 + 2E.
• It has been stated that electrolyses in acid chloride solutions may be
carried out in the presence of toluene; the chlorine liberated during electro-
lysis is said to react preferentially with the toluene and therefore does not
attack the platinum electrodes.
Gravimetric Analys~s 607
The acid concentration of the solution must not be too great,
otherwise the deposition of the copper may be incomplete or the
deposit will not adhere satisfactorily to the cathode. The beneficial
effect of the nitrate ion is due to its depolarising action at the
cathode:
N0 3- + IOH+ + 810 = NH4+ + 3Hp.
The reduction potential of the nitrate ion is lower than the discharge
potential of the hydrogen, and therefore hydrogen is not liberated
in the free state. The nitric acid must be free from nitrous acid as
the nitrite ion hinders complete deposition and introduces other
complications. The nitrous acid may be removed by boiling the
nitric acid before adding it, or by the addition of urea to the solution:
2HNO z + CO(NHz)z = 4N z + CO2 + 3HzO.
The err?r due to nitrous acid is increased by the presence of a large
amount of iron; 'iron is reduced by the current to the ferrous state,
whereupon the nitric acid is reduced. This error may be 'minimised
by the proper regulation of the PH and by the addition of ammonium
nitrate instead of nitric acid, or, best, by the removal of the iron
prim to the electrolysis.
The solution should be free from the following, which either
interfere or lead to an unsatisfactory deposit: silver, mercury, bis-
muth, selenium, tellurium, arsenic, antimony, tin, molybdenum,
gold and tile platinum metals, thiocyanate, chloride, oxidising agents
such as oxides of nitrogen, or excessive amounts of ferric nitrate or
nitric acid. The electrolytic deposit should be salmon-pink in
colour, silky in texture, and adherent. If it is dark, the presence of
foreign elements and/or oxidation is indicated. Spongy or coarsely
crystalline deposits are likely to yield to high results; they arise
from the use of too high current densities, or improp~r acidity and
absence of nitrate ion.
Procedure. The solution (100 ml.) may contain 0·2 to 0·3
g. of Cu (1). Add cautiously 2 ml. of concentrated sulphuric
acid, 1 to 1·5 ml. of concentrated nitric acid (the latter should
be freshly bailee.. in order to remove oxides of nitrogen), and
transfer to, unless already present in, the electrolysis vessel.
For simplicity of description, it will be assumed that the
cathode is of the gauze form (Fig. 4-10 a) and the anode i~
either a thick platinum wire spiral (Fig. 4-8 f) or a gauze cylin-
der (Fig. 4-10 b); the student should be able to readily adapt
the procedure to other types of electrode if he reads Section
IV. 77 carefully. Clean the platinum gauze cathode by heating
it in 1 : I nitric acid, and washing it thoroughly with distilled
water. Dry the electrode at 100° to 110°C, cool in a desic-
cator, and weigh. Handle the electrode by the stem and not
by the gauze, since a trace of grease may cause a non-
adherent deposit of copper. Arrange the circuit as shown in
Fig. 4-4, or suitably adapted according to the source of current
(if accumulators or storage batteries supplying not more than
12 volts are employed, one large resistance will suffice), but
do not connect the source of current (2). Be sure that the
catltode is connected to the negative terminal and the anode
to the positive terminal. If a rotating anode is to be used
subsequently, make c~rtain that it can rotate without coming
into contact with the cathode at any point. Place the elec-
trolysis vessel in position (e.g., a beaker resting on a wooden
block or upon a stand), and adjust the height so that the
electrodes extend nearly to the bottom of the beaker and the
cathode is 80 to 90 per cent immersed in the solution. Cover
the electrolysis vessel (beaker) with a split clock glass (Fig.
4-14), and with all the resistance in the circuit, so that only a
small current will flow, close the circuit and proceed as (A)
or (B)t depending upon whether or not stirring is employed.
(A). Slow electrolysis.. without stirring. With an appl'ied
potential of 2 to 2·5 volts, adjust the resistance until a current
of approximately 0·5 amp. flows, as indicated by the ammeter
in the circuit: electrolyse, preferably overnight (3). Rinse
off the split clock glasses and test for complete deposition
after the blue colour of the solution has disappeared. This is
best done by adding more water to raise the level of -the
electrolyte (say, by 0·5 cm.) and continuing the electrolysis
for 30 minutes. If at the end of this time no copper deposit
has appeared on the freshly immersed surface of the cathode,
it may be assumed that the deposition is complete. If a
deposit does form, continue the passage of the current as long
as may be judged necessary, and again test as before. If the
solution is not to be used for a further estimation, a drop or
two of it may be removed and tested with sodium acetate and
hydrogen sulphide water or with potassium ferro cyanide
solution.
(B) Rapid electrolysis, with stirring. Start the motor
driving the anode (or, less satisfactorily, the glass propeller)
and adjust its speed (300 to 800 r.p.m.) so that the solution is
vigorously stirred without there being any danger of mechani-
cal loss of liquid. Use a voltage across the terminals of the
cell of 3 to 4 volts, and adjust the resistance so that the
current is 2 to 4 amps. Continue the electrolysis until the
blue colour of the solution has entirely disappeared '(usually
somewhat less than I hour), reduce the current to 0·5 to 1
amp., and test for completeness of deposition by rinsing the
split clock glasses, raising the level of the liquid by about 0·5
cm. by the addition of distilled water, and continuing the
electrolysis for 15 to 20 minutes. If no copper plates out on
the fresh surface of the cathode, electrolysis may be regarded
as complete.
When electrolysis has been shown to be complete, the
subsequent procedure is the same whether slow or rapid
electrolysis has been employed .. Two methods may be
employed: (i) is used only when the residual electrolyte is not
required for further determination, and (ii) is of universal
application.
(i) Siphon off the liquid (for example, with a glass siphon
provided ~ith a stop cock). Whilst siphoning the heavier
liquid from the bottom of the beaker, add water from the top
to keep the level of the liquid in the electrolysis vessel nearly
constant. Continue this process until the ammeter needle
drops practically to zero. Then rinse the cathode with a
little pure alcohol or with A.R. acetone (4), and dry it for 5 to
10 minutes at 100° to 110°C. Weigh when cold.
(ii) Lower the beaker very slowly, or raise the electrodes, and
at the same time direct a continuous stream of distilled
water from a wash bottle or supply bottle against the upper
edge of the cathode. This washing must be done imme-
diately the cathode is removed out Qf the solution and the.
circuit must not be broken during the process. When the
eathode has been thoroughly washed, break the circuit, dip
the cathode into a beaker of distilled water, and then rinse it
with alcohol or A.R. acetone (4). Dry at 100° to 110°C for
5 to 10 minutes, and weigh when cold.
From the increase in weight of the cathode, calculate the
copper content of the solution. After the cathode has been
weighed, it should be cleaned with nitric acid as described in
Section lV, 77, and reweighed ;' the loss il,1 weight will serve as
a check.
Notes, 1. Larger quantities of copper may be present, particularly if rapid
electrolysis is employed; the quantity given is, however, convenient for
instructional purposes. For .practice in the determination, the student may
prepare the solution either by weighing out accurately about 1. g. of A.R,
copper sulphate pentahydrate or by dissolving about 0·25 g., ac;_curately'
weighed, of A.R. copper in 1 : 1 nitric acid. boiling to remove nitrous fumes.
just neutralisirtg with ammonia solution, and then just acidifying with dilute
sulphuric acid and diluting to 100 mi.
2. This is, -of course, unnecessary u' a commercial .. electro-analyser" or
electrolysis unit is employed.
X
610 Quan#tative Inorgam:c Analysis
3. If conditions do not allow this practice. use a current of 1·5 to 3·0 'amps. :
deposition is usually complete in 2 to 4 hours. This procedure is. however.
less satisfactory.
4;. This is best effected by completely immersing the cathode in a beaker
containing pure alcohol or acetone.
IV,79. LEAD
Discussion. Lead is deposited quantitatively as the dioxide at the
anode in the presence of a high concentration of nitric acid (15 ml.
of concentrated acid per 100 ml. of the electrolyte). With a gauze
anode of the usual size (Fig. 4-10 a) aboutO'1 g. ofleadis the maximum
that can be firmly deposited. With a platinum dish (matted or
sand blasted internal surface) as anode, quantities up to 2 g. may
be readily deposited; this method is therefore frequently employed
when large quantities of lead are to be determined in one operation.
It is difficult to remove all the water from the electrode by drying
at low temperatures. For a temperature of 120°C, the conversion
factor 0·8640 instead of the theoretical conversion factor of 0·8662
is employed. However, by dryiD§' at 220°C for about an hour, the
theoretical factor may be employed for quantities up to 0·1 g. For
quantities of o·r to 0'3 g., the factor 0·8650 has been recommended,-
whilst for larger weights 0·8635 has been suggested (L. Classen, 1927).
A useful method is to dissolve -the dioxide, without drying. in
standard oxalic acid solution, and to titrate the excess with standard
potassium permanganate solution.
The following interfere in this determination: mercury, arsenic,
antimony. tin, selenium, tellurium, phosphorus. chromium, chloride,
iodide, silver, bismuth and manganese (the last three metals tend to
fonn peroxides at the anode).
Procedure. For a platinum gauze electrode (Fig. 4-10 a),
the solution (100 mq should contain not more than 0·1 g.
of Pb as lead nitrate, 15 ml. of concentrated nitric acid, and
none of the interfering elements mentioned above. Some-
what greater quantities may be deposited with the aid of
Sand's electrodes (Figs. 4-11 and 4-12). Larger quantities of
lead are generally determined, as already pointed out, with a
platinum (Classen) dish. The present details apply, hQwever,
chiefly to the gauze electrode, .but can be readily adapted to
the use of a platinum dish.
Heat tl].e anode to 220°C in, an electric oven for 1 hour,
allow to cool in a desiccator, and weigh. Connect up the
cell with the positive terminal of the ..source of current on to
.the gauze imode or platinum dish, and the negative terminal
to the wire or gauze cathode (I). Adjust the current to 0·05
to 0·1 amp. at 2 volts with the aid of the-rheostat, and allow
t~e electrolysis to proceed overnight. Test for completeness
of deposition by adding about 20 ml. of distilled water and
continuing the electrolysis for 15 minutes; if no darkening
of the freshly-covered anode surface occurs, deposition is
complete. When the electrolysis is complete, either lower the
beaker from the electrodes or slowly raise the electrodes out
of the solution without interrupting the current, and at the
same time'rinse the electrodes very thoroughly with a jet of
water from a wash bottle or from a storage bottle. Then
disconnect the source of current, wash the anode with alcohol
or acetone, and dry. at 220°C; cool in a desiccator and weigh.
The determination may be carried out more rapidly by one
of the following methods.. In these the outer electrode is the
anode.
(i) With a current of 1·5 to 2·0 amps. at 2 volts at the ordin-
ary temperature: Electrolysis is complete in about 1·5
hours.
(ii) With a current of about 1·5 amps. at 2 volts at 50° to
65°C. About 0·1 g. of lead dioxide can be deposited in 12 to
18 minutes; the solution should contain 20 per cent of its
volume of cOl,lcentrated nitric acid. '
(iii) With 2 volts and an initial current of 0·5 amp., which is
subsequently raised to 5 amps. 0·1 Gram of lead dioxide
may be deposited in 6to 8 minutes at room temperature: this
time is still further reduced by working at about 60°C.
The deposited dioxide is removed from the electrode, after
weighing as above, by immersion in warm 1 : 1 nitric acid to
which a little pure hydrogen peroxide has been added. As an
additional check the solution may be evaporated with sul-
phuric acid, and the lead determined as •sulphate (Section
,IV. 14A).
Note 1. Full experimental details of the general technique are given under
Copper (Section IV. 78). and will therefore not be repeated in the description
of this and succeeding determination,
IV, 80. CADMIUM

Discussion. Cadmium is best determined from a faintly alkaline
solution containing only enough potassium cyanide to keep the
cadmium in solution, i.e., containing the complex K2[Cd(CN)J
(Procedure A). Less accurate results, but sufficiently precise for
most routine analyses, may be obtained in very dilute sulphuric
acid solution; higher current densities are permissible when a
rotating electrode is employed (Procedure B).
Procedure A. The solution should contain about 0·4 g. of
Cd as the sulphate, acetate, or, less desirably, the nitrate.
Add a drop of phenolphthalein, followed by ca. O'IN sodium
612 Quantitative Chemical Analysis
or potassium hydroxide until a permanent pink colour is just
obtained. Then add a solution of pure (e.g., A.R.) potassium
or sodium cyanide dropwise and with constant stirring until
the precipitated cadmium hydroxide just dissolves. A large
excess"of alkali cyanide should be avoided. Dilute to 100 to
150 ml., and electrolyse the cold solution, preferably with a
weighed platinum gauze cathode, and a current of 0·5 to 0'7
amp. at 4·8 to 5·0 volts (1). At the end of 6 hours, increase
the current to 1·0 to 1·2 amps., and continue the electrolysis
for another hour. Wash the split .clock glass and the sides
of the beaker with about 20 ml. of water, and continue the
electrolysis for 15 minutes. If the newly.exposed surfa€e of
the cathode remains bright, thus indicating that the deposition
of the cadmium is complete (2), remove the electrolyte from
the electrodes, rinse immediately with water, stop the current,
and rinse the_ cathode with alcobol 01: A.R. acetone. Dry
at 100°C, cool, and weigh. Test the residual electrolyte for
cadmium by any of the recognised tests.
, The determination may be carried out more rapidly by
using a rotating cathode (600 to 800 r.p.m.) with a current of
1·5 to 2·0 amps. at 2·7 to 3·0 volts. 0·2 Gram of Cd may
thus be deposited in 30 minutes. ,
Procedure B. The cadmium should be present in ~he solu-
tion (100 'ml.) as sulphate; nitrates and chlorides, must be
absent. The maximum concentration of free sulphuric. acid
is 0·5N, and 5 g . .of potassium hydrogen sulphate is added.
Electrolyse at room temperature with 0·1 tf> 0·2 amp. at 2'4
to 2·8 volts; after 3 hours, increase to 0·5 amp. until electro-
lysis is complete. 0·3 Gram of Cd is thus deposited jn -3 to ,4
hours.
Alternatively, use a rotating electrode with 1·5 to 0·7
amps. at 2·7 volts. 0·2 Gram of Cd may thus be deposit'M in
20 minutes.
Notes. l. If left overnight, use a current of 0·2 to 0,3 amp. at 2,8 to 3'2
volts.
2. It is sometimes difficult to detect the deposition of the bright cadmium
on the platinum 5urface.- This difficulty is readily overcome by heavily plating
the platinum electrode with copper or silver first, and then proceeding with
the electrolysis in the usual manner. An added advantage is that the removal
of the cadmium after the electrolysis is rendered more facile; also, if the temper-
ature of drying should accidentally exceed 100°C, there is..tittle danger of harm-
ing the platinum electrode.
IV, 81. NICKEL

Discussion. The most satisfactory method for the determination
of nickel by electrolysis is from strongly ammoniacal solution con-
Gravimetric' Analysis 613
taining ammonium sulphate. The nickel is present largely as the
complex ion :
[Ni(NH3)6)++ ~ Ni++ + 6NH 3,
and is deposited as a dense adherent layer, bright silver-grey in
colour, upon the cathode. The nickel should be present preferably
as the sulphate, but the' chloride may also be employed; the nitrate
introduces difficulties, and should be absent. The metals of the
hydrogen sulphide group, zinc, vanadium, tungsten, molybdenum,
chromates and ferrous salts should be absent; cobalt is deposited
with the nickel. SmaU amounts of the precipitated hydroxides, such
as that of iron, are not occluded by the deposit to any appreciable
extent; as a general rule, however, they are best removed first.
The electrolysis should not be unduly prolonged, because there is a
slight tendency towards solution of the anode, with subsequent
deposition of platinum upon the cathode.
Procedure. The solution should contain not more than about
'0·2 g. of Ni as sulph.ate cir chloride. Add 5 to 10 g. of ammo-
nium sulphate, 30 to 40 m!. of concentrated ammonia solution,
and dilute to 150 ml. Electrolyse using 1 amp. at 3 to 4
volts; about 0"15 g. of Ni is deposited in 2 hours (1). Test for
completeness of precipitation in the usual way by adding
about 20 ml. of water, and continuing the electrolysis for 15
to 20 minutes; no nickel sho-qld be deposited on the freshly
immersed surface. (2). Alternatively, the dimethylglyoxime
test, suitably modified, may be applied. Wash, 'dry and weigh
the cathode as in the estimation of Copper (Section IV t 78).
The determination ,may be carried out more rapidly by
using a rotating electrode with 4 amps. at 3-4 volts. 0·1
Gram of nickel is thus deposited in about 10 minutes.
The nickel is best removed fr6m the electrode by means of
dilute mineral acids; concentrated nitric acid should not be
employed because of the danger of inducing passivity.
'Note. 1. With II- current of 0·2 to 0·3 ampere, the deposition is generally
complete in 12 to 15 hours. This current is used if the electrolysis can be
allowed to proceed overnight.
2~ This test is not always'satisfactory, for deposits of dickel are sometimes
obtained v.:tich can hardly be distinguished from the platinum of the electrode
(compare Caclmhnn, Section IV, 80, Note 2).
IV,82. COBA.Ur
Discussion. The determination of cobalt foilows along the same
lines as given under Nickel (Section IV, 81). The addition of a little
hydrazine sulphate improves the character of the cathode plate by
reducing the tendency to form an oxide deposit at the a~ode.
614 Quantitative Inorganic Analysi$
IV, 83. SILVER
Discussion. Silver may be determined by electrolysis in nitric,
a.mmoniacal or cyanide solutions. In cyanide solution the silver is
present largely as the complex ion:
[Ag(CN)J- ~ Ag+ + 2CN-;
an excellent plate is obtained, and separation from other elements
(e.g., copper and lead) may be effected. The cyanide method will
be described.
The disadvantage of the electrolytic method is that so many other
elements are also deposited, either wholly or in part. that a number
of preliminary separations are usually required before it can be
applied. For this reason, it is not widely employed.
Procedure. The silver (ca. 0·2 g.) s?ould be present in
solution as the nitrate. Add pure (e.g., A.R.) potassium
cyanide until the precipitate 01 silver cyanide is dissolved, and
then add an excess such that about 3 g. of potassium cyanide
is present in the solution. Dilute to _100 to 120 ml. Elec~
trolyse with 0·2 to 0·5 amp. at 3'7 to 4·8 volts at 20° to 30°C;
about 0·1 g. of Ag is deposited in 3 hours. Alternatively,
electrolyse with a rotating electrode with 1 to 0·5 amp. at
2·5 to 3·2 volts; 0·2 g. of Ag is deposited in 20 to 25 minutes.
Completeness of deposition is tested for by transferring a few
drops of the electrolyte to a test tube, acidifying with a little
nitric acid, boiling off the hydrocyanic acid (caution': poison~
ous), rendering ammoniacal, and adding a few drops of
ammonium sulphide solution; no brown precipitate should
be obtained. The determiI).ation is <?ompleted as under
Copper (Section IV, 78).
NQte. If insoluble silver salts are to be analysed, e.g., chloride,
bromide, iodide and oxalate. these may be dissolved directly in the
potassium cyanide ·solution.
IV, 84. ZINC
Discussion. Zinc is quantitatively deposited from an alkaline
zincate solution containing excess of alkali. The proportion of alkali
is not of great importance provided that it ~xceeds a minimum of
10 molecules of alkali hydroxide to 1 moI'ecule of zinc salt: it is
unnecessary to employ a larger excess of alkali than is re_quired to
keep t1;1e solution clear during the whole of the electrolysis. Am-
monium salts, if ·presen.t, must be decomposed by the addition of
excess of caustic alkali and boiling. Chlorides and nitrates inter-
fere; they must be removed, if present. by evaporation to fuming
with sulphuric acid. .
Zinc should not be directly deposited on a platinum surface as it is
difficult to remove it completely (it tends to leave black stains) and
upon heating may alloy with the noble metal. The platinum elec-
trode should be plated with copper or, preferably, with silver.
Alternatively, a silver gauze cathode (as Fig 4-10 a) may be used.
Cathodes of nickel have also been used, and are quite satisfactory.
A useful application of the electrolytic method is in the analysis of
alloys containing aluminium and zinc.
Procedure. For practice in this method, weigh out accur-
ately about 0·.3 g. of A.R. zinc oxide, dissolve it in the mini-
mum voll1tne of dilute sulphuric acid, add a solution of 7 g.
of A.R. ~odium hydroxide in water, stir until the solution is
clear, dilute to 125 ml., and electrolyse using a copper or silver
plated platinum cathode or a silver cathode. For slow elec-
trolysis, use a current of 0·4 to 0·8 amp. at S·8 to 4·2 volts:
For rapid electrolysis (rotating electrode), the current may be
increased up to 5 amps. The metal is more rapidly deposited
at about 60°C. When electrolysis is complete, remove the
solution without interrupting the current, wash thoroughly
with water, then witli alcohol or acetone, dry at 60 0 to 70°C,
and weigh. Calculate the percentage of Zn in the sample.
IV, 85. Electrolytic separation and determination of copper
and nickel.-
Discussion. This determination has been included at this stage in
order to indicate the use of the electrolytic method in the separation
and determina~ion of metals. The theory is discussed in Section
I, 69. There are a number of alloys, whicq include Monel metal,
certain coinage alloys (e.g., American), and " cupro-nickel," which
are composed principally of copper and nickel, together with small
amounts of iron and manganese and not more than traces of other ele-
ments. These are suitable for this exercise in ~Iectrolytic separation.
The copper is determined in strongly acid solution at a potential
not exceeding 4 volts (above this potential nickel may plate out).
The solution is· evaporated to fuming in order to remove
excess of nitric acid, the iron present is precipitated with ammonia
solution, and the nickel deposited from the filtrate after the addition
'of a large excess of ammonia solution.
The separation of lead and copper by electrolysis is described in
Section IV, f5i (analysis of brass). -
Silver· and copper in silver-copper alloys (silver solder, sterling
silver, and certain coinage alloys) may be separated and determined
electrolytically by dissolving the sample in 1 : 1 or 1 : 2 nitric acid,
removing the excess of nitric acid by evaporation, and then estimat-
ing the silver after the addition of excess of pure potassium cyanide
at a voltage not exceeding 1·5 (Section IV, 83). The residual solu-
tion is acidified with nitric acid, the hydrocyanic acid boiled off, and
the coppel" deposited as detailed in Section IV, 78. Alternatively,
the silver may be deposited from ammoniacal solution (5 mL of free
concentrated ;:unmQnia solution per 100 ml, of solution). Most of
616 Quantitative Chemical A'nalysis
the silver is deposited with a current of 0·4 amp., a little hydrogen
peroxide is added to oxidise cuprous salts and to redissolve any
particles of silver not plated on the, electrodes, and the deposition.
completed at 0·2 amp. It is essential that the solution be vigorously
stirred during the whole of the estimation. The copper is deter-
mined in the residual electrolyte after acidifying with nitric acid and
diluting.
Procedure (analysis of a copper-nickel allOY). Weigh out
accur!ltely about 0·5 g. of the clean alloy (I) into a 150 ml.
beaker, which should be suitable as an electrolytic vessel.
Add a mixture of 10 ml. of water, 1 ml. of concentrated
sulghuric acid and 2 ml. of concentrated nitric acid to dissolve
the alloy. When solution is complete, boil off the oxides of
nitrogen, and dilute to 100 ml. The solution is now ready for
the deposition of copper.
Copper. Proceed as directed in Section IV, 78, and employ
either the slow or rapid method of electrolysis. Wash the
copper deposit thoroughly with water, and keep the solution
for the estimation of iron and nickel.
Iron. Evaporate the solution and washings from which the
copper has been removed orr a low temperature hot plate as
far as possible, and then heat at a higher temperature until
fumes of sulphur trioxide appear. Cool the residue, and
carefully add water until the volume is about 25 ml. ' Preci-
pitate the small quantity of iron that is now present in the
ferric state by adding to the warm solution about 10 ml. of
I : I ammonia solution in excess. Filter through a small
quantitative filter paper, and collect the filtrate in a 150 ml.
electrolytic beaker (A). Wash the precipitate three times
with water. Place the original beaker under ..the ;filter;
dissolve the precipitate in a little hot 1 : 5 sulphuric acid and
wash the paper with water. Precipitate the iron again with
the same large excess of 1 : 1 ammonia solution, and filter
through the same paper. Wash the precipitate; and collect
the filtrate and washings in beaker (A) containing the filtrate
and washings from the first precipitation. Ignite the ferric
hydroxide to ferric oxide, and weigh it (Section IV, 8).
Nickel. Add 15 ml. of ·concentrated ammonia solution to
the ammoniacal nickel solution, and dilute to 100 to 120 ml.
Carry out the determination as detailed in Section IV, 81.
Calculate tht> percentage of copper, iron and nickel in the
alloy.
Note. 1. Either Hidsua!c's .. Cupro-Xickcl, No. I II" (onc of the Analysed
Samples for Students) or th" Bureau of Analys!'d Samples" Cupra-Nickel
, ;\ , " (one of the British Chemica! standards) is suitable for this estimation.
IV, 86. SIMPLE GRAVIMETRIC SEPARATIONS
Discussion. The author has found it a good practice in his own
classes for students to undertake simple gravimetric separations
bef.ore embarking upon the analysis of complex materials. This
scheme may not commend itself to all teachers, chiefly on the ground
that tp.e mixtures are artificial and are therefore not likely to be
encountered ~s such in practice; however, the analysis of simple
artificial mixtures is a valuable preliminary training before proceed-
ing to the analysis of complex materials.
Many separations have been briefly indicated in the text; a few
are given in detail. With his knowledge of the determination of the
component elements, the student should experience no difficulty in
carrying out the undermentioned separations and estimations.
Procedure. A. Iron and aluminium. Method 1. Dissolve
the mixture provided (1) in water or in dilute hydrochloric
acid in a Pyrex beaker, and dilute to 75 m!. (2). Heat to
abol,lt 70°C, and add an excess (8 to IO-fold) of O'2N_sodium
hydroxide (A.R.) slowly and with constant stirring (3). The
excess of alkali will dissolve the precipitated aluminium
hydroxide as sodium aluminate. Dilute to 150 m!. with
boiling water, boil for several minutes, and filter. Wash
the beaker, precipitate and filter with hot water until the
washings are at most faintly alkaline. Reserve the filtrate
and washings (i). Place the original beaker under the filter,
dissolve the precipitated ferric hydroxide by pouring warm
2N hydrochloric acid upon it (4), and then reprecipitate the
iron with ~mmonia solution (Section IV, 8), or by any of the
methorls given in Section IV, 31. Ignite and weigh as Fe20 a.
Determine the aluminium in the filtrate (i) either by acidi-
fying with dilute hydrochloric acid, adding seve.al grams of
ammonium chloride, and precipitating witli dilute ammonia
solution (Section IV, 9), or by just acidifying with dilute
hydrochloric acid and precipitating with sodium thiosulphate
solution (Section IV, 28A)'; in both cases the precipitate is
19nited to and weighed as A1 20 3 • Alternatively, the oxine
method (Section IV, 28B) may be used.
Notes. 1. A suitable mixture for practice may be prepared by intimately
mixing known weights of A.R. iron alum and A.R. ammonium alum.
2. If ferrous iron is present, oxidise it with a little concentrated nitric acid.
3. Improved results may be obtained by the following method of precipita-
tion. Nearly neutralise the dilute acid solution of iron and aluminium with
aqueous sodium hydroxide. heat to boiling. and add the solution slowly and
with constant stirring to 75 m!. of hot 5 per 'cent sodium hydroxide (A.R.)
solution. Boil the mixture for 2 to 3 minutes with gentle stirring. and filter.
4. It is difficult to completely free the precip'itate from sodium hydroxide.
etc.; this is most simply achieved by reprecipltation.
x· 6i7
Method 2. This procedure depends upon the precipitation of
iron by ammonium sulphide in ammonium tartrate solution:
aluminium (also chromium, vanadium, titanium, zirconium
and beryllium) is not precipitated under these conditions.
Dissolve the mixture provided in water or in dilute hydro-
chloric acid in a conical flask, anti dilute to 75 to 100 mI.
Add 4 times as much pure tartaric acid as the dissolved
. metals would weigh as oxides, render alkaline with ammonia
solution (the solution should remain clMr if enough tartrate
is present), acidify with 1 : 1 hydrochloric acid, and add 2 ml.
in excess. Saturate. with hydrogen sulphide to red.uce the
iron (also to take advantage of the fact that ferrous sulphide
is filtered more easily than ferric sulphide). Add a slight
excess of dilule ammonia solution, introduce more hydrogen
sulphide, and allo~ the precipitate to settle in the closed
flask. Filter off the precipitated fero£ous sulphide, wash
thoroughly with water containing a little ammonium sulphide
and ammonium tartrate, and keep the filtrate and washings
(i). Dissolve the precipitate in I : I hydrochloric acid. Alter-
natively, place the precipitate and paper in a beaker of suit-
able size, cover with a clock glass, and cautiously dissolve the
sulphide in 1 : 1 hydrochloric acid: heat until all the black
particles have disappeared, and shred the paper ,by vigorou~
·stirring with a policeman. Oxidise the iron by the addition
of 1 ml. of concentrated nitric acid, boil, dilute, and precipitate
the iron as in Method 1. Weigh, as Fe2 0 a.
To determine the aluminium, the excess of tartrate must be
destroyed. Add cautiously 10 to 12 ml. of concentrated
sulphuric acid to the filtrate (i), evaporate in a large platinum
dish, and heat carefully until the solution begins to show signs
of charring. Allow to cool, cover with a clock glass, and add
5 mI. of concentrated, better of fuming, nitric acid. When
the violent action has ceased, heat gradually until the organic
.matter is fully oxidised; repeat the treatment with nitric
acid, if necessary. Cool, dissolve the residue in water, and
precipitate the aluminium with ammonia solution as in
Section IV, 9. Weigh as Alps.
M efhod 3. This method illustrates the use of cupferron in
separating iron (and also titanium, zirconium and vanadium)
from aluminium (and also from chromium, manganese and
nickel). Precipitation is made in strongly acid solution.
Dissolve the mixture provided in water or in dilute sul-
phuric acid. Dilute to 150 mI., and add cautiously 15 m!. of'
concentrated sulphuric acid. Cool to 10°C, and then intro-
duce a freshly-prepared, cold 6 per cent aqueous solution of
cupferron slowly ana' with constant stirring until an excess*
is present. A reddish-brown flocculent precipitate is pro-
duced; precipitation'is complete when a white, finely crys-
talline precipitate of cupferron appears. Add a Whatman
accelera~or (or one-third of an ashless tablet), ·allow to settle
for 2 to 3 minutes, and filter with gentle suction through a
quantitative filter paper supported on a filter cone (Whatman,
-hardened, No. 52). Wash tbe precipitate with cold 10 per cent
by volum'e sulphuric or hydrochloric acid, then twice with 5N
ammonia solutibI1 to remove excess of the.reagent, and finally
once with water. Reserve the filtrate and washings (i).
Complete the determination asin Section IV, 31B. Weigh as
Fe 20 a•
To estimate the aluminium, concentrate the filtrate a,nd
washings (i) to 25 ml., add 20 to 25 m!. of concentrated nitric
acid, and evaporate slowly to ftlmes of sulphuric acid.
Treat with more nitric aci9- if ·organic matter is still pr~sent.
Finally, evaporate just to dryness, dissolve the residue in
water, and complete the determination as in Method 2.
For other methods of separation, see Section IV,95, CI and
C2.
B. Iron and Chromium. Method 1. Dissolve the mixture
(1) in water or in dilute hydrochloric acid, almost neutralise
with A.R. sodium hydroxide solution, and heat to boiling.
;pour this solution slowly and with constant stirring into. 100
ml. of a hot, 5 per cent A.R. sodium hydroxide solution, boil
for 2 to 3 minutes, and cool to the lab~ratory temperattlre.
Treat :with bromine water until the solution is yellow (alter-
natively, add a few ml. of lOO-volume hydrogen peroxide) ;~
boil to decompose the excess of hypobromite (or of hydrogen
peroxide). The chromium wilt now be ill soluti?n as C]iro-
mate. Add a Whatman accelerator (or one-third of an
ashless tablet), allow the precipitate to settle, and filter
through.a quantitative filter paper. Wash the precipitate
with hot water, and reserve the filtrate and washings (i),
It is advisable to dissolve the. precipitate in 2N hydrochloric
acid, and to reprecipitate'the ferric hydroxide as in A, Method
1. Weigh as FezO a. •
If sulphate is absent, the chromium may be determined as
barium chromate. Render the filtrate (i) acid 'with acetic
acid,. and add an excess of barium acetate solution. Complete
• About 33 per cent excess, i.e., 0·83 g. of cupferron for 0·1 g. of Fe.
620 !
Quantitative norganic Analysis
the estimation as -in Section IV, 309. Weigh as -BaCrO,.
If sulphate is present, acidify the filtrate (i) with hydro-
chloric acid, reduce the solution by the addition of 15 ml. of
alcohol, and boil. Precipitate the chromium as chromic
hydroxide as in Section IV, 9 or Section IV, 30A. Weigh as
Cr2 0 3 •
Note. 1. A suitable mixture for practice may be prepared by intimately
mixiI;lg known weights of A.R. iron alum and A.R. chrome alum.
Method 2. Here the chromium ·is oxidised in acid solution
by perchloric acid to chromate (1).
Weigh out a suitable quantity of the mixture into a conical
flask, add 5 ml. of watet and 5 ml. of A.R. 60 to 70 per cent
perchloric acid (DANGER; see Section IV,. 70), -and place a
short funnel in the mouth of the flask. Heat in the fume
cupboard until the solution has evaporated to half its original
volume, and has assumed a deep orange colour for at least 5
minutes; oxidation is then complete. Allow to cool, add 40
to 5Q ml. of water, and boil for 2 to 3 minutes to drive off any
free chlorine which may be present (test with starch-iodide
paper). Transfer quantitatively to a 250 ml. beaker, and
precipitate the iron by the addition of ammonia solution in
the usual manner (Section IV, 8). Keep_ the filtrate and
washings (i). The ferric hydroxide should be dissolved in 2N
hydrochloric acid and reprecipitated. Weigh as Fe 2 0 a.
Determine the chromate in the filtrate and washings (i)
as in Method 1 or by the following volumetric process. Acidify
with 1 : 1 hydrochloric acid, transfer quantitatively to a 350
or .500 ~l. glass-stoppered bottle, and treat with ·30 ml. of
10 per cent potassium iodide solution and 5 ml. of concentrated
.hydrochloric acid. Add 50 m!. of water, and allow to stand
for 1 minute. Titrate the liberated iodine with standard
O'IN sodium thiosulphate:. add starch solution towards the
end, and continue the titration until the green colour appears.
1 Ml. N Na 2S 20 S ;::: 0·02533 g. Cr 20 3 •
The above procedure may be applied aiso to the separation
of aluminium and chromium.
Note. 1. The following reactions probably occur:
Cr.O. + 3HCIO, = 2CrO. + 3HCIO.;
3HCIO. = HCIO, + CI. + 20.;
or CrlO a + -2HCIO. = 2CrO. + CI I + 20 2 +
F!P.
C. Manganese and zinc. This separation depends upon the
solubility of manganese quinaldinate in dilute acetic ·add,
the corresponding zinc complex being insoluble.
Cravimetric Analysis 621
Dissolve the mixture (1) in 200 ml. of water, heat, and add 2
ml. of glacial acetic acid. Treat the boiling solution slowly
and with stirring with a slight excess of sodium quinaldinate
reagent (Section IV, 340). Allow the precipitate to settle,
filter it through a weighed sintered glass or porous porcelain
crucible, and wash successively with hot 1: 40 acetic acid and
hot water. Reserve the filtrate and washings (i). Dry the
precipitate at 125°C to constant weight, and weigh as
Zn(C 1oH,ON 2)2,H 20.
Determine the manganese in the filtrate (i) as the pyro-
phosphate (Section IV, 35).
Note. 1. A suitable mixture for practice may be prepared by intimately
mixirlg known weights of A.R. manganese sulphate with A.R. zinc sulphate or
pure' zinc ammonium sulphate (Section IV, 3U).
D. Nickel and zinc. Weigh out accurately a suitable
quantity of the mixture (1) containing not more than 0·2 g.
of the total metals, dissolve it in 200 ml. of water, and add 1
ml. of concentrated hydrochloric or sulphuric acid. Heat the
solution to boiling, add a large excess of the dimethylglyoxime
reagent (Section I. 62A) followed by a concentrated solution of
ammonium acetate until a precipitate appears and about 2 g.
in excess. Digest for 39 to 60 minutes, and filter. Com-
plete the estimation as in Section IV, UA. Weigh as
Ni(C,H 7 0 zN z)z. Reserve the filtrate and washings (i).
For the estimation of zinc it is necessary to destroy the
excess of dimethylglyoxime. To do this, acidify the filtrate
(i) with concentrated hydrochloric acid, and add 5 ml. of the
concentrated acid in excess. Then introduce 25 ml. of con-
centrated nitric acid, and evaporate on a hot plate in the fume
cupboard to dryness or to fu~es of sulphur trioxide. Allow
to cool, add 5 ml. of concentrated hydrochloric acid, warm
the mixture, and dilute to 50 ml. Complete the determina-
tion as in Section IV, 34A (compare Section IV,89 (1))-
r~member that the proportions of the reagents apply to about
0·1. g. of Zn. Weigh as both the ammonium phosphate and
the pyrophosphate.
Note. 1. A suitable mixture for ptactice may be prepared by intimately
mixing known weights of pure (e.g., A.R.) nickel ammonium sulphate and A.R.
zinc sulphate. •
E. Calcium ~nd magnesium. The calcium is separated by
double precipitati'on as calcium oxalate, whilst magnesium is
determined by double precipitation as the ammonium phos-
phate. An alternative method which permits of the quantita-
tive separation of the calcium in one operation is to precipi-
tate it as calcium molybdate: magnesium is then determined
in the filtrate as magnesium ammonium. phosphate. For full
experimental details, see Section IV, ~4D,.E and Section IV,
94H respectively.
A suitable mixture for practice may be prepared from A.R.
calCium carbonate and A.R. magnes'ium sulphate.
F. Calcium and barium. This separation is based upon the
difference in solubilities of calcium and barium chromates in
acetic acid.
Weigh out a suitable quantity of the mixture (I), dissolve
it in the minimum quantity of dilyte hydrochloric .acid in a
covered beaker, boil to expel carbon dioxide, if present, and
allow to cool. Neutralise with dilute ammonia solution,
acidify with a few drops of acetic acid, and dilute to 50 to 100
ml. Introduce 10 ml. of a 30 per cent solution of A.R.
a.:mmonium a~etate (this should be faintly ammoniacal), heat
to boiling, and add dropwise and with constant stirring an
excess Qf 10 per cent A.R. amlllonium bichromate solution
(5 to 10 m!.). Allow to cool for 1 hour, decant the clear yellow
liquid through a quantitative filter paper (or thr6ugh a porous
porcelain crucible), and wash the precipitate by decantation
with 2 per cent ammonium acetate solution until the filtrate
is no . longer per.ceptibly coloured (ca. 100 ml. of the wash
liquid). Complete washing is unnecessary at this sta~e (2).
Reserve the filtrate and washings {i). Tntns£er the precIpitate
by a jet of water from the wash bottle to the beake,r containing
the bulk of the precipitate, place the beaker under the-funnel,
and dissolve the residual precipitate in warm 2N nitric acid,
and wash the filter with hot water until free from acid, Add
more dilute nitric acid, if necessary, until the precipitate
dissolves. Treat with 10 ml. of 30 per cent ammonium acetate
solution, heat the liq~id to boiling and stir steadily all the
while, and allow to cool slowly during I hour. (The super-
natant liquid should be y~llow; if this is not the case, add 'a
few m!. of ammonium bichromate solution.) Filter the
precipitate through a weighed silica Gooch or porous porcelain
crucible; wash first by decantation, then on the filter with
2 per cent ammonium acetate solution, and finally with cold
water until the filtrate gives only a slight opalescence with 2
drops of O·IN silver nitrate. H.eserve the filtrate and wash-
ings (ii). Dry the crucible and precipitate at 500 0 to 600°C
or within a larger crucible at a dull red heat to constant weight.
Weigh as BaCr0 4 • •
Treat the combined filtrates and washings (i and ii) with a
little dilute nitric acid, evaporat~ to a small volume, and
precipitate the calciul}l as carbonate by the addition of
ammonia and A.R. ammonium carbonate solutions. Filter
off the precipitate of calcium carbonate, wash it with a little
hot water, dissolve it in the minimum volume of 1: 1 hydro-
cploric acid, and precipitate the calcium as oxalate (Section
IV, 10) and weigh either as CaC0 3 or as CaO, or preCipitate as
molybdate (Section IV, 94H) and weigh as CaMo0 4 •
Notes. 1. For practice in this analysis, an intimate mixture of A.R. barium
chloride and A.R. calcium carbonate may be employed.
2. The subsequent reprecipitation is unnecessary where results of only
moderate accuracy are required. For the separation of barium and strontium
this method may also be used, but reprecipitation is essential.
G. Calcium and strontium. The method (R. N. Shreve,
C. H. Watkins and J. c. Browning, 19a9) is based upon the
fact that calcium nitrate is soluble in anhydrous acetone
whilst strontium nitrate is almost insoluble (0'02 per cent at
25°C).
Weigh out a suitable quantity of the mixture (1), dissolve
it in the minimum volume of dilute hydrochloric acid in a
.. covered beaker, boil to expel carbon dioxide (if present), and
allow to cool. Warm to 50°C, and precipitate the two ele-.
ments as carbonates by the addition- Qf excess of ammonia
and A.R. ammonium carbonate solutions. Digest the
precjpitate on a steam bath or on a hot plate for 10 minutes,
and a~low to cool. 'Filter off the precipitate, and wash it with
a little water. Dissolve the precipitate in dilute nitric acid,
and ·evapor.ate the solution to dryness on a water bath.
Allow to cool, add 25 ml. of anhydrous acetone, and cover
wit,h a clock glass; leave the solvent In contact with the
mixed nitrates for 1 hour with occasional agitation. Transfer
the residual strontium nitrate to a weighed filter crucible,
wash with more acetone, dry at 100°C, and weigh as Sr(NO a)2'
Determine the calcium in the filtrate as CaS04 (the details
are similar to those given under H below).
Note. 1. For practice in this analysis, an Intimate mixture of known weights
of A.R. strontium chloride or carbonate and A.R. calcium carbonate may be
used.
H. Calcium, strontium, and barium. Although this analysis
is hardly suitable for elementary students of gravimetric
analysis, it is given here as an _illustration of the separation
and determination'of three similar elements.
Precipitate the three elements as carbonates by, the addition
of excess of ammonia and A.R. ammonium carbonate solu-
tions. Filter off the precipitate, and'wash it with a little hot
624 Quantitative I norganic A naZy sis
water. Dissolve the presipitate in dilute nitric acid, and
evaporate the solution to dryness on the water bath; heat
~or 30 minutes in an air bath (Section U, 12F or Fig. 2-13)
at 135° to 140°C, and for 30 minutes at 180°C. Extract the
powdered residue twice with 2·5 mI. portions of absolute ethyl
alcohol, and once with 5 ml. of dry isobutyl alcohol (1).
Dissolve the residue i~ water, evaporate to dryness on the
water bath, and heat as above. E"tract again with two 5 m}.
portions of isobutyl alcohol. Evaporate the alcoholic filtrates
cautiously in a weighed crucible (preferably of the Main-
Smith type), and convert the r~sidual calci,um nitrate into
calcium sulphate as follows. Add 1 or 2 drops of water, then
5 times the weight of finely-powdered A.R. ammonium sul-
phate, and dry at 100° to 110°C for 30 minutes. _ Place the
crucible inside a larger porcelain crucible provided with an
asbe~tos supporting ring (Fig. 60) and heat with a small
flame until fuming commences, then increase the size of the.
flame. Continue the heating for 1(2 hour after all fuming has
ceased. Allow the crucible to cool, and weigh as CaSO,.
(Too strong heating-above dull. redness-causes loss .of·
sulphur trioxide.)
Dissolve the residue containing the nitrates of strontium
and barium in water or in dilute acetic acid, and determine the
barium as chromate as' detailed in F. Precipitat'e the· stron-
tium in the filtrate as the carbonate (the experimental pro-
cedure is similar to that given 'for calcium in F), filter, and
'wash with a little hot water; dissolve the pre~ipitate in a
small volume of dilute. hydrochloric acid and wash the filter
with a little hot water. Add slowly a lO-fold excess of dilute
sulphuric acid and then a volume of alcohol equal to that of
the solution. Complete the determination as in Section
IV, 44A. and weigh as SrS0 4• (Here also too strong heat-
ing-above dull redness-results in loss of sulphur trioxide.)
Note. 1. Commercial isobutyl alcohol is refluxed over' dry baryta or over a
little potassium hydroxide pellets, and then fractionated to separate the
fraction of boiling point 107° to 10Soe. Two drops should not give a cloudi-
ness when mixed with 5 ml. of pure paraffin.
I. Sodium and potassium. The potassium in the mixture
of chlorides is determined as potassium perchlorate and
therefrom the equivalent amolmt of potassium chloride is
computed. The sodium is obtained 1Jy difference. For
experimental cletails, see Section IV, 48, B, B'.
A suitable mixture for practice may be prepared by inti-
mately mixing A.R. sodium and potassium cWQrides.
ANALYSES OF COMPLEX MATERIALS*
IV,87. ANALYSIS OF BRASS

Discussion. The chief components of the alloy brass are copper
and zinc, but small quantities of tin, lead and iron, and sometimes
very small quantities of other clements as well, are usually present.
The percentages of the important constituents rarely fall outside the
following values: '
Cu Zn Sn Pb Fe
50-90% 20-40% 0-6% 0-2% 0-1 %.
The analysis involves, therefore, the separation and determination
of tin, lead, copper, iron, and zinc.
Outline of procedure. The sample is treated with nitric acid to
precipitate the hydrated tin oxide Sn02 ,nH20 and to bring copper,
zinc, iron and lead into solution: the insoluble stannic acid is filtered
off,'ignited and weighed as stannic oxide, Sn02.t The filtrate from
the tin precipitate is evaporated with an excess of sulphuric acid to
fumes of sulphur trioxide in order to expel nitric acid. The residual
liquid is treated' with water, and the insoluble lead sulphate is
filtered off, and weighed as PbS0 4 after ignition at about 500°C
(see, however, Section IV,l4A), The copper in the filtrate from the
lead sulphate is determined electrolytically by deposition on a
platinum cathode, or is estimated as cuprous thiocyanate, Cu 2(CNS)2'
The solution from which the copper has been separated is treated
with bromine water or with nitric acid to oxidise iron to the ferric
state, and this metal is separated from the zinc by double precipita-
tion with ammonia solution; the ferric hydroxide is ignited and
weighed as Fe20 a. Finally, the zinc is precipitated as zinc ammon-
ium phosphate, and weighed either as zinc ammonium phosphate
ZnNH 4P04 or, after ignition, as zinc pyrophosphate, Zn 2P 2 0 7 •
An alternative procedure for the electrolytic determination of
* Space does not permit the inclusion of the experimental details for more
than a few typical complex materials. For an excellent and authoritativ~
account of many others the reader, is referred to Lundell, Hofmann and
Bright. Chemical Analysis of Iron and Steel (J. Wiley, 1931). This treatise
contains details for inter alia iron ores; mangav,ese \)res; chrome ores;
fluor spar; refractories of high alumina content an.d bauxite: slags; molding
sands; coal and coke: spiegeJei,sen., ferrom"ngalJ.ese and manganese metal;
ferrosilicon and refined silicon; ferrochromium and chromium metal; fen:o-
vanadium; ferromolybdenum and calcium molybdate; ferrotungsten ;
1errocobalt and cobalt metal; ferrotitaniuID; ferrozirconium; ferrotanta-
lum; and ferroboron. A standard reference work dealing with the analysis
of complh materials is W. D. Scott-No H. Furma.n, Standard Methods of
Chemical Analysis, Vols. I and II (D. Van Nostrand Co.; The Technical Press
Ltd., 1939).
t Tin in brass. bronze. etc., may also be determined with phenylarsonic
acid (see Section IV, 220).
625
626 Quant#ative Inorganic Analysis
lead by deposition as lead dioxide at the anode is described. If any
antimony is present, it will be weighed with the tin; any aluminium
present will separate with the iron.
Procedure. If the sample, which should be in the form of
drillings or turnings, is not quite clean and dry, wash it with a
little ether, and dry it by exposure to air on a watch glass for
a short time.
(a) Tin. Weigh out accurately about 1 g. of the brass (1)
into a 150 ml. beaker, add 10 ml. of water followed by' 15 ml.
of concentrated nitric acid; keep the beaker covered with a
clock glass or, better, a Fisher" speedyvap II. beaker cover.
When the vigorous reaction is over, evaporate on a water
bath to a volume of 5 to 10 mI., but not to dryness. The
evaporation should reql}ire at least an hour to insure that all
the stannic acid has precipitated (2); longer digestion does.
no harm. Dilute to 50 ml., and heat on the water bath for 10
to 20 minutes in order that all soluble salts may dissolve.
Then add one Whatman " accelerator," and stir. Filter hot
through a small Whatman No.' 42 or No. 542 paper, and
collect the filtrate (i) in a 150 ml. beaker. (If ijle filtrate is not
clear, refi.lter it, and add more macerated filter paper if the
precipitate persists in running through.) Wash the precipi-
tate at least 10 times with hot I per cent nitric acid solution.
Fold up the filter paper and its contents into a weighed porce-
lain crucible, dry and char the filter at as low a temperature
as possible, and finally ignite the residue for 30 minutes over
a Mcker or similar type burner under good oxidising condi-
tions (compare Section IV, 22A). Allow to cool in a desiccator
and weigh. Repeat the ignition until constant weight is
obtained.
If the residue is coloured, foreign oxides are present. The amount of
stannic oxide present may be determined by the excellent ammonium iodide
method (Section IV, 22A) or, less rapidly, by the foJl()wing procedure. Mix
the weighed residue with S times its weight of a mixture of equal parts of pure
sodium carbonate and pure sulphur,_cover the crucible with a lid, and heat
with a low flame in the fume cupboard until the sulphur ceases to burn from
under the crucible covel'. Allow to cool, and then place the crucible and its
cover in a beaker containing sufficient water to cover both, and warm to dis-
solve the melt. (During this process sodium polysuJphide is formed, whi<;:h
reacts with the stannic oxide to give the soluble sodium thiostannate, and
forms insoluble sulphides with the other metals. The latter sulphides, par-
ticularly that of copper, are appreciably soluble in polysulphide solutions, but
not in solutions of normal sulphides; hence it is necessary to decompose the
polysulphide by the addition of sufficient sodium sulphite to form the colour-
less sulphide, for example:
Na.S._ + Na.SO. = Na.S.0 3 + Na.S.)
Gravimetrt'c Analysis 627
After the melt has dissolved. remove the crucible and lid from the solution.
and wash them free from the adhering solution. Then add excess of sodium
sulphite solution. and boil for 10 to 15 minutes. Filter off the residue. which
may contain the sulphides of copper. lead and iron. through a quantitative
filter paper. wash with 0 ·IM colourless ammonium sulphide solution. and ignite
in a weighed porcelain crucible. The weight of the ignited residue. which
may consist of CuO. PbO and Fe,Oa. is subtracted from the weight of impure
stannic oxide. The ignited residue should be dissolved in 1 to 2 ml. of 1 : 1
nitric acid. a~d added to the main filtrate containing the copper. lead and zinc.
Notes. ,. The determination should be conducted in duplicate. Ridsdale's
.. Brass. No. 5b" (one of the Analysed Samples for Students) is suitable for
this estimation.
2. If no precipita.te has appeared at the end of 1 hour. no appreciable quan-
tity of tin is present. •
(b) Copper and lead. There are three alternative methods
~f dealing with the main filtrate and washings from the tin
separation (i). In Method A the lead is separated and
weighed as lead sulphate, and the copper is separated and
weighed as cuprous thiocyanate. In Method B the lead is
sep!lrated as lead sulphate, and the copper determined electro-
lytically. In Method C both the lead and copper are deter-
mined electrolytically, the lead as lead dioxide at the anode
and the copper as metallic copper at the cathode.
Method A. Add to the.main filfrate (i) from the metastannic
acid or, if tin was absent, to the nitric acid solution of the
brass 3 to 4 m!. of concentrated sulphuric acid, and evaporate
the solution carefully {e.g., on a low temperature hot plat¥)
until white fumes are evolved (3). Allow the solution to cool,
wash down the sides of the beaker (or the casserole, if em-
ployed) with 10 to 20 mi. of water, and again evaporate to
white fumes in order to insure the expulsion of all the nitric
acid. Cool, add 25 mi. of water cautiously, heat nearly to
boiling, and stir for several minutes until all soluble salts pass
into solution. Then add 50 m!. of water, stir thoroughly, and
allow to stand for at least 1 hour. Filter off the 'precipitate
through a weighed silica Gooch or porous porcelain crucible,
and wash thoroughly with cold 1 : 50 sulphuric acid. Reserve
the filtrate and washings (ii). Ignite the lead sulphate to
constant weight at 500 to 600°C (see Section IV, l4A in
0
which other methods for weighing the PbSO, are described).

Calculate the percentage of lead.
Evaporate the filtrate (ii) from the precipitation of lead
SUlphate to about 50 mI., almost neutralise the free acid with
1 : 1 ammonia solution, and precipitate the copper as cuprous
thiocyanate (Se_ftion IV, 19A). Res~rve the filtrate and
washings (iii) for the d~termination of iron and zinc. Weigh
as Cu 2 (CNSh after drying at 110° to 120°C.
628 Quantitative Inorganic, Analysis
Calculate the percentage of copper.
Method B. ·Determine the lead as lead sulphate as described
under Method A. Concentrate the filtrate and washings (ii)
from the lead sulphate to about 75 mI., add 4 mI. of concen-
trated ammonia solution to reduce the acidity, and then 2 ml.
of concentrated nitric acid. Estimate the copper electroly-
tically; use the slow or~rapid method described in Section
IV,78.
The combined solution and ·washings (iii') after the depo-
sition of copper is kept for the determination of iron and zinc.
It may contain a small amount of copper. This may be
recovered by passing hydrogen sulphide into the solution for
a few minutes. If only white sulphur is formed, copper is
absent. If a small dark precipitate is formed, filter it through
a small filter paper, wash with 1 per cent ammonium nitrate
solution, and ignite to the oxide in a weigheq. porcelain crucible.
Add the weight of copper contained in the cupric oxide to the
weight obtained 011 the electrode. Evaporate the filtrate to
remove hydrogen sulphide (iii), a~d use it for the determina-
tion of iron and zinc. .
Method C. Evaporate the filtql.te ang washings from the
metastarinic acid {i) to 50 to 75 mI. Transfer quantitatively
~o an electrolytic beaker of 150 ml. capacity, add 1·5 ml. of
concentrated nitric acid for each 10 mI. of solution, and then
0·5 ml. of concentrated sulphuric acid (to make the deposit
more adherent). Introduce two clean and weighed electrodes
into the solution. These may consist of a platinum gauze
electrode (Fig. 4-10 a), which is connected to the positive
terminal of the source of current, and a heavy platinum wire
spiral (Fig. 4-8 f) or a platinum gauze electrode (Fig. 4--10 b),
which is connected to the negative terminal and therefore
acts as cathode. Alternatively, Sand's electrodes (Fig. 4-11
or Fig. 4-12), similarly connected, may be used. When all is
ready, bring the 150 mI. beaker containing the· solution
beneath the electrodes, and adjust the height so that the
electrodes extend nearly to the bottom, and the anode is at
least three-quarters immersed. Switch on the current, and
adjust the resistance to give 1·5 to 2 amps. at 2 volts (4).
Electrolyse for 1·5 hours (5). Test for completeness of
deposition by adding about 20 m!. of water, and observing
whether any deposit is formed on the freshly:.immersed gauze
or stem of the anode when the electrolysis is continue.d for 15
to 20_ minutes. When the electrolysis is complete, lower the
beaker gently or slowly raise the electrodes out of the solution
whilst rinsing the electrodes thoroughly by a stream of water
from a wash bottle, and take care that all the washings fall
without loss into the electrolytic beaker; the current should
not be interrupted during the washing. When the electrodes
are out of the solution and have been washed, turn off tqe
current. Rinse the anode well.in a beaker of water, then with
acetone to remove water. Dry the anode at 220°C, cool in a
desiccator, and weigh (compare Section IV, 79). Calculate
the percen'tage of lead. Clean the electrode as described in
Section IV, 77.
Some cbpper may have been deposited on the cathode. If
this is the case, place the cathode in the electrolysed solution,
and allow it to stand therein until the copper has been dis-
solved off. Then remove the cathode, rinse it with water, arid
reserve the solut~on for the determination of copper.
For the estimation of copper, it is necessary first to remove
the large excess of nitric acid present in the solution. Aad 3
ml. of concentrated sulphuric acid r and evaporate the solution
until fumes of sulphur trioxide appear. Cool, dilute with
water to 75 to 100 mL, add 2 ml. of concentrated nit,ric acid,
and determine the copper as described in, Section IV, 78.
Calculate the percentage of copper. Reserve the liquid and
washings (iii).
Notes. 3. Nitric acid increases the solubility of lead sulphate, and hence
must be completely removed.
4. Excellent results are obtained by electrolysing overnight with 0·5 amp.
at 2 volts.
5. Any of the electrodes shown in Figs. 4-7, 4-8, or 4-9 may be used.
(C) Iron. The exact procedure will depend upon whether
the copper has been separated electrolytically (Method B) or
as cuprous thiocyanate (Method A). In the former case the
iron will be present in the ferrous state. In the latter case,
the excess of thiocyanate and sulphite must be removed by
evaporation with nitric and hydrochloric acids, which also
relnoves ammonium salts.
Method D (copper separated electro~cally). Heat the
solution (iii) from the electrolysis to boiling, add 3 to 4 ml. of
saturated bromine water or 1 ml. of concentrated nitric acid,
and boil for 2 to 3 minutes. Add concentrated ammonia
solution until a precipitate is obtained, and then 10 ml. in
excess. Allow the precipitate to settle, filter it through a
small filter paper into a 400 ml. beaker, and wash the precipi- !
tate 3 times with 1 : 100 ammonia solution. Replace the

•
400 m!. beaker by the beaker in which the precipitation was
made. Pour' 2(). ml. of hOt 1 : 3 hydrochloric acid through
the filter to dissolve the precipitate, wash the filter first with
1 : 100 hydrochloric acid and then with water to remove acid.
Precipitate the iron in the filtrate as before, using 5 ml.
excess of concentrated ammonia solution. Filter the preci-
pitate through the same filter paper into the first filtrat~,
and wash the precipitate thoroughly with hot water .containing
a little ammonium nitrate and a drop or two of concentrated
ammonia solution. Reserve the combined filtrates (iv) for
the determination of zinc. Ignite the precipitate to ferric
oxide (Section IV, 8) and weigh. Calculate the percentage of
iron ..
• Method E (copper separated as cuprous thiocyanate). To the
filtrate from the thiocyanate precipitation (iii) add 35 ml. of
concentrated" nitric acid and 15 ml. of concentrated hydro::.
chloric acid, and' evaponrte the solution nearly to dryness.
Treat the cold residue with 3 m!. of concentrated hydro-
chloric acid, warm to dissolve the residue, and dilute to 50 to
75 ml. If there is a precipitate (silica), filter it off on a small
filter paper, and wash thoroughly \yith I : _rOO hydrochloric
acid. Determine the iron in the solution.as in Method D.
(d) Zinc. This constituent of brass is usually determined by
precipitation as zinc ammonium phosphate, which is subse-
quently ignited to and weighed as zinc pyrophosphate.
Excessive amormts of ammonium salts must be removed by
oxidising them with hydrochloric and nitric acids. If the
thiocyanate method for copper was used, the ammonium
salts were removed before the iron was separated. ,If,
however, the lead or copper was determined electrolytically,
the solution will contain a considerab.le quantity of ammonium
salts. Remove them as follows. To the filtrate from the
ferric hydroxide (iv) add 25 ml. each of concentrated hydro-
chloric acid and concentrated nitric acid, and evaporate the
solution to dryness or to fumes of sulphur 'trioxide. Tr~at
the cold residue with 3 nil. of concentrated hydrochloric acid,
warm for a few minutes, dilute with water to 50 mI., and filter
off any silica which may have separated, collecting the filtrate ..
in a 400 ml. beaker.
Transfer the filtrate quantitatively to a 250 ml. volumetric
flask, make up to-the mark, and shake well. Remove 100 ml.
of the solution,'" add 5 g. of ammonium chloride, 10 ml. of
* This \olume contains.about 0'1 g. of Zn.
G1'avimetric Analysis 631
2N sodium acetate, and dilute to 140 ml. Heat on the water
bath. and, when hot, treat with 10 ml. of 10 per cent A.R.
ammonium phosphate (NH4hHPO, added slowly from a
pipette or burette. Complete the determination as described
in Section IV, MA. Weigh the zinc first as ZnNH4P0 4 after
drying at 105°C for about 1 hour, and then as Zn ZP Z0 7 after
ignition. Calculate the percentage of zinc. Finally, sum up
the percentages of tin, lead, copper, iron and zinc.
Note. If tne es.timation of zinc has been spoilt because of some
accident, it is desirable in a repeat analysis to avoid the time re-
quired for the separation of the other metals. The following method
is applicable in the presence of large amounts of copper and small
amounts of tin and lead, and is based upon the use of the tartrate
ion to prevent the precipitation of the phosphates of these metals.
Dissolve 1 g., accurately weighed, of brass in 30 ml. of I : I nitric
acid, and eX4ctly.neutralise the solution with 1 : 5 ammonia solution,
using litmus paper as indicator. Heat the solution to about 90°C,
add 25 g. of ammonium tartrate Qissolved in 75 ml. of water, heat
again to 90°C, and add slowly and with constant stirring a solution
of 7 g. of A.R. (NH4)2HPO, in 35 ml. of water. Keep the solution at
80 0 to 90°C for 1 hour, decant the supernatant liquid through a
quantitative filter paper, transfer th~ precipitate to the paper, and
wash it with 1 per cent ammonium tartrate solution. Dissolve the
precipitate in 1 : 3 hydrochloric acid, and wash the filter first with
I per cent hydrochloric acid and finally with water. Neutralise the
filter and washings very carefully with I : 5 ammonia solution, using
litmus paper, add 15 g. of ammonium tartrate dissolved in 5() ml.
of water, and reprecipitate the zinc with ammonium phosphate.·
Filter the precipitate through a weighed porous filtering crucible or
through a silica Gooch crucible, and wash it with water. Ignite the
zinc ammonium phosphate to the pyrophosphate in the usual way
(Section IV, 34A).
IV,88. ANALYSIS OF BRONZE

Discussion. The chief components of the alloy bronze are copper
(60-97%) and tin (1-35%), but small amounts of lead, iron and zinc
may be present. Gunmetal is cOl1lposed nominally of 90% of copper
and 10 % of tin, but small amounts of zinc (up to 2 %), lead and iron
together with traces of phosphorus are always present. Phosphor
bronze is similar in general composition to gunmetal, but contains
also 0·2 to 1·5 % of phosphorus and, sometimes, small amounts .of
antimony. English bronze coins contains about 95% of copper, 4%
of tin, and 1% of zinc. Aluminium bronze usually consists of
about 90% of copper and 10% of alUlpinium, but may also contain
• For the most accurate results, it is desirable to dissolve the precipitate and
reprecipitate again.
small amounts of iron and other metals. Manganese bronze* contains
55 to 60% of copper, 30 to 40% of zinc, 0 to 2% of tin, 0 to 1·5% of
lead, I to 2% of iron, 0 to 3% of aluminium, and about 1% of
manganese. .
Only the analysis of gun metal and of phosphor bronze will be
described here. The student should employ his own ingenuity in
devising schemes for the analysis of the other bronzes, and should
check up the procedure by reference to any of the standard text
books on metallurgical analysis (see Appendix, Section A, 3).
For the analysis of gunmetal and of phosphor bronze, the experi-
mental details are similar to those already given under Brass (Section
IV, 87) for the determination of copper, lead, iron and zinc. The
precipitated metastannic acid will carry down considerable quanti-
ties of lead, copper and iron, and also phosphorus, if present, pre-
sumably as stannic phosphate. It is therefore essential to deter-
mine the amount of tin present in the impure precipi~ate. Three
methods may be used.
Method A. Transfer the precipitate of metastannic acid and
macerated filter paper to a beaker, 'add 10 mI. of ammonium hydro-
gen sulphide (NH4HS) solution, prepared by saturating concentrated
ammonia solution with hydrogen sulphide, and warm until the
precipitate largely dissolves: The sulphides of copper, lead and
iron will remain unattacked. Filter t.hrough a hardened quantita-
tive filter paper, and wash with dilute ammonium hydrogen sulphide
water. Dissolve the sulphides on the paper in a small volume of
1 : 1 nitric acid and a little bromine water, and add the solution to
the main filtrate from the metastannic acid. Heat the ammonium
sulphide solution of tin to boiling until a precipitate commences to
form, add 10 to 15 m!. of concentrated nitric acid, and evaporate to
dryness. Add 50 m!. of 1 : 10 nitric acid, heat to boiling, add one
Whatman accelerator, filter through a Whatman No. 42 or No. 542
paper, and wash the precipitate with 1 per cent nitric acid. Cqm-
plete the estimatiDn as d~scribed in Section IV, 87 (a), and weigh
the tin as stannic oxide.
Method B. Ignite the impure metastannic acid, and weigh as
Sn02 + P20S + foreign oxides. Heat the impure weighed oxides
with a large excess of ammonium iodide at 425° to 475°C, treat the
residue with nitric acid, ignite and weigh. By subtracting the
latter weight from that previously obtained, the weight of
(Sn0 2 + P 20S)t is secured. A separate portion is analysed for
phosphorus (see below) and the weight of P20S is subtracted from
that of (Sn0 2 + P 20 S)' If phosphorus-is absent, the method is
of course much simplified.
• This is sometimes termed high tensile brass.
t See Section, IV, 22A. The phosphorus and stannic tin are volatilised as
the iodides. The residue in the crpcible may be dissolved in hot concentrated
hydrochloric acid, and added to the main filtrate for the determination of the
other metals.
Gravimetric A natysis 633
Method C. Transfer the precipitate of me~astannic acid' and
macerated filter paper, preferably dried at 100°C, to a Kjeldahl
flask (see Section ill, 20), treat with 15 m!. of concentrated sulphuric
acid and 5 g. of anhydrous sodium sulphate, and boil gently until
the organic matter is destroyed. Dilute cautiously, transfer to a
flask fitted with carbon dioxide tubes (Section m, 116, Procedure B),
add 2 to 3 g. of pure lead, excess of concentrated hydrochloric afid,
and boil gently in an attposphere of carbon dioxide. The quadriva-
lent tin is reduced by tIle lead to the bivalent state:
SnCI, + Pb + 2HCI = SnCl2 + PbC1 2 + 2HCl.
Allow to cool in a stream of carbon dioxide, and then titrate the
stannous' tin with standard iodine solution (Section m, 116,
Procedttre B). From the volume of standard iodine solution
required, calculate the percentage of tin.
For the determination of the other metals, treat another sample of
the alloy with 1 : 1 nitric acid, etc., and dissolve the impure meta-
stannic acid in a mixture of concentrated sulphuric acid and sodium
sulphate as above. Transfer the mixture quantitatively to a beaker,
treat cautiously with excess of sodium hydroxide solution followed
by excess of sodium sulphide, (pld digest for some time. The tin
thus passes into solution as sodium thiostannate Na 2SnS 3 , and the
sulphides of the other metals (e.g., iron, coppe~, zinc, etc.) are pre-
cipitated. Allow to cool, dilute somewhat, and filter. Dissolve the
residue in hot 1 : 1 nitric acid, and add the solution to the main
filtrate for the estimation of the other metals.
Analysis of gunmetai. Procedure. Use 0·5 to 1 g., accurately
weighed, of the clean dry alloy (I), and proceed as detailed
in Section IV. 87 for Brass. The tin should be determined by
any of the three methods given above, but preferably by
Method B (2). Finally, sum up the percentages of tin, copper, -
lead, iron and zinc in the sample.
Notes. 1. Ridsdale's" Gunmetal, No. 6b" (one of the Analysed Samples for
Students) is suitable for this analysts. .
2. Perhaps the simplest method for the determination of tin is with phenyl.
arsonic acid (Section IV, 22C.)
Analysis of phosphor bronze. Procedure. Use 0·5 to 1 g.,
accurately weighed of the alloy (3), and follow the method
used for gunmetal (4). Determine the phosphorus in a fresh
sample as follows.
Weigh out accurately about 2 g. of the sample into .a 300
to 350 m!. conical flask, and dissolve it in 20 to 25 m!. of aqua
regia. When all the alloy has dissol:ved, add 10 m!. of·water,
and digest at 90°C for 10 minutes. Dilute to 50 to 60 mI., and
adjust the temperature to 40° to 50°C. Add 100 ml. of the
ammonium molybdate. reagent, previously heated to 40° to
634 Quantitative Inorganic A na~'Ysis
50°C, slowly and with constant stirring. Complete the
estimation as described in Section lV, 65B. Dissolve the
precipitated ammonium phosphomolybdate in 5N ammonia
solution, and precipitate,. the phosphate with magnesia mixture
as magnesium ammonium phosphate. Ignite the latter
and weigh as magnesium pyrophosphate. From the weight
of the latter, calculate the percentage of phosphorus in the
sample. •
Finally, sum up the percentage of tin, copper, lead, iron,
zi~c and phosphorus in the phosphor bronze.
Notes. 3. Ridsdale's" Phosphor Bronze. No.7" (one of the Analysed Sam-
ples for Students) or the }3ureau of Analysed Samples " Bronze' A , .. (one
of the British Chemical Standards) is suitable for this analysis.
4. If the ammonium iodide method for tin is used. it must be remembered
that both tin and antimony (the latter is often present in samples of phosphor
bronze) are volatilised. The tin alone may be determined by the phcnylar-
sonic acid method (Section IV, 220). or by Method C (above). and hence the
antimony content estimated by difference. •
IV, 89. ANALYSIS OF GERMAN SILVER (NICKEL SILVER)

Discussion. The chief components of German silver are copper
(46-63%), zinc (20.:_34%). and nickel (6-25%) ; ,small quantities of
iron, lead and tin may also be present. The procedure described
allows for the determination of all these elemenfs and is given in
outline only since most of the essential details have already been
given under Brass (Section IV, 87).
Procedure. Weigh out accurately about 0·5 g. of German
silver in the form 'of clean drillings or turnings into a tall 150
to 200 mI. beaker. Cover the beaker and add slowly 10 ml.
of concentrated nitric acid. When the reaction is almost
over, warm'the mixture on a hot plate until the alloy has
completely dissolved. Add 10 ml. of water, and allow the'
solution to digest for 10 to 15 minutes.
(a) Tin. If a white residue is present, it is metastannic
acid. This should be separated and determined according to
the procedure for tin in Brass (Section lV, 87 (a) ).
(b) Lead. Add 2 to 3 m!. of concentrated sulphuric acid to
the filtrate from the metastannic acid or, if tin is absent, to
the original solution. Evaporate the solution to a pasty, mass
and fumes of sulphur trioxide (not to dryness) to remove all
the nitric acid. To the cold residue add 30 ml. of water
cautiously, stir thoroughly, and allow to stand for at least
an hour in order that the precipitation of lead sulphate may
be complete. From this point follow the experimenta,l details
~iven for lead in Method A 'under Brass.
(c) Copper. The filtrate from the lead sulphate, or the
original solution if both tin and lead are absent, is employed
for the electrolytic determination of copper. Determine the
copper electrolytically; use either the slow or rapid method
described in Section IV, ~8 (see Method B in Section IV. 87 (b)).
(d) IroD:. Add 1 to 2 m!. of hydrogen peroxide to the solution
from the copper determination in order to reoxidise the iron
to the ferric state. Add a large excess of concentrated
ammonia solution, and heat to boiling. Filter off the pre-
cipitated ferric hydroxide, and wash it with hot water. Dis-
solve the precipitate in the minimum volume of hot 1 : I
hydrochloric acid, and receive the acid solution in a small
beaker. .Pour in enough I : 2 ammonia solution to reprecipi-
tate the ferric hydroxide, heat to boiling, filter and' wash the
precipitate. Add the filtrate and washings to the main
filtrate contail,ling the zinc and nickel. Ignite. the precipitate,
and weigh the ferric oxide.
(e) Nickel. Add .hydrochloric actd to the combined filtrates
from the determination of iron until I1eutral (to litmus paper),
and add 5 m!. in excess. Add ammonia solution slowly until
the solution is faintly ammoniacal. Ammonium chloride is
formed· by this neutralisation process and this must be present
in order to hold the 'zinc in solution; the same result is
achieved by adding 5 g. of ammonium chloride to the faintly
ammoniacal solution. Fpr each 1)·1 g. of niCkel that is expected
to be present, add to the hot solution 0·5 g. of dimethyl-
glyoxime (1) dissolved in rectified spirit, and make certain
that. the solution is slightly ammoniacal (add more dilute
ammonia solution, if necessary). Allow the beaker to stand
on the steam bath for 15 to 20 minutes, and test the solution
for complete precipitation after the red. precipitate has settled
out. complete the determination as described in Section
IV, 12A.
Note. 1. Too large an excess of dimethylglyoxime must be avoided as
otherwise it, too, may be precipitated.
(1). Zinc. Acidify the filtrate from the determination of
nickel with concentrated hydrochloric acid, and add 5 ml. of
the concentrated acid in excess. Then add 25 ml. of concen-
trated nit:t:ic.acid, and evaporate on a hot plate in th~ fume
cupboard either to dryness or to fumes of sulphur trioxide if
much sulphuric acid is present. This evaporation removes the
orgai!ic matter and the large amount of ammonium salts
present in solution. Treat the cold residue with 5 ml. of
Quantitative Inorganic A n:a1ysis
concentrated hydrochloric acid, warm the mixture for a few
minutes; dilute to 50 ml., and, when the salts have-dissolved,
filter the solution if not perfectly clear. Complete the deter-
mination as described for zinc in Brass (Section IV, 87 (d)).
Weigh (he zinc as both the ammonium phosphate and as the
pyrophosphate_
IV, 90. ANALYSIS OF SOLDER

Discussion. Ordinary solder is an alloy of approximately 3 parts
of lead and 2 parts of tin, but it usually contairis also from 1 to 3
per cent of antimony as impurity. \Vhen such an alloy is disin-
tegrated by nitric acid, the insoluble residue consists of metastannic
acid and antimonic acid, together with some coprecipitated lead
+
compound. The ignited residue will consist of Sn02 Sb20 4 PbO.+
The amount of lead present in the residue is most simply deter-
mined by the ammonium iodide volatilisation method (Section IV,
22A); the tin and antimony are volatilised as the iodides, and the
lead remains behind as a mixture of oxyiodide and oxide. The
residue in the crucible is treated with 2 to 3 ml. of concentrated
nitric acid (a Main-Smith crucible is very suitable for this process),
evaporated to dr~ess on the hot plate, the nitrate cautiously de-
composed, and finally ignited to oxide at a low red heat. The loss
in weight gives (Sn0 2 + Sb20 4 ). The tin may be determined as in
Method C, Section IV, 88 (antimony does not ir.terfere, with this
process), and the anti~ony obtained by difference. The residual
PbO in the crucible may be dissolved in a little concentrated hydro-
chloric or nitric acid, and added to the main filtrate from the
metastannic acid. The lead is determined as sulphate (Section IV.
14A) or as chromate (Section IV:14B); in the latfer case, about 12
g. of sodium acetate are added to replace the free nitric acid by
acetic acid.
An alternative procedure is to transfer the impure metastannic
acid precipitate and filter paper, preferably dried at 100 0 e, to a
Kjeldahl flask, add a mixture of 20 ml. of concentrated sulphuric
acid and 5 g. of anhydrous sodium sulphate, and the whole boiled
gently until the organic matter is destroyed. The carbonaceous
matter from the filter paper reduces the antimony to Sbm . The
mixture is evaporated to fumes of sulphur trioxide, allowed to cool,
cautiously transferred quantitatively to a conical flask with the aid
of about 50 ml. of water, concentrated hydrochloric-acid, cooled to
woe, and titrated with standard O'IN potassium permanganate (see
Section m, 117 : _
5Sb2 (SO,)a + H 2S04 + 4KMn04 + 4H 20
= lOH aSb0 4 + 2K 2S04 + 4MnS0 4 + 10H 2S0 4 -
The percentage of antimony can then be calculated. Excess of

concentrated hydrochloric acid is then added, followed by 2 to 3 g.
Gravimetric A na_lysis 637
of pure lead, and the mixture boiled gently in an atmosphere of
carbon dioxide (see Method C in Section IV. 88). The quadrivalent
tin i~ reduced by the lead to the bivalent state, and the antimony is
precipitated as the metal. The stannous tin is titrated with stan-
dard O·IN iodine, and from the volume of the latter required the
percentage of tin may be calculated (Section m,1l7).
The procedure is much simplified if antimony is absent, as irr the
case of a pure solder or a pure lead-tin alloy. Such alloys are
available commercially, and are useful for practice in the quantita-
tive separation and determination of lead and tin. Here the amount
of lead oxide present in the impure stannic oxide maybe determined
either by the ammonium iodide process or by the so$lium carbonate-
sulphur fusion process (Section IV, 87 (a)). The lead is determined
in the filtrate from the impure-metastannic acid, to which the small
amount of recovered lead has been added, as the sulphate.
Analys"'ls of pure solder or a pure lead-tin alloy. Procedure.
Weigh out accurately about 0·5 g. of the finely-divided alloy
(turnings, drillings, etc.)(I) into a 150 to 200 ml. beaker,
and cover with a clock glass or Fisher" speedyvap " cover.
Add 5 ml. of water, followed by 15 ml. of concentrated nitric
acid: keep the beaker covered during the ensuing violent
reaction. When the vigorous reaction is over, evaporate on'a
water bath to a volume of about 5 ml., but not to dryness.
Dilute to 50 mI., heat on a water bath for 10 to 15 minutes,
then add one Whatman accelerator (or half of an ashless
"tablet), and stir. Filter through a Whatman No. 42 or No.
542 paper, and collect the filtrate in a 250 to 400 ffil. beaker.
Wash the precipitate at least 10 times with 1 per cent nitric
acid. Keep the filtrate .and washings for the determination
of lead. Fold up the filter paper and precipitate and place it
in a weighed porcelain crucible (best in a Main-Smith
crucible), and complete the estimation,as described forTin in·
Section lV, 87 (a). Weigh the impure stannic oxide. Add
to the latter about 15 times its weight of powdered ammonium
iodide, and mix the two in the crucible intimately by means
of a small spatula. Heat the crucible in an electric crucible
or muffle furnace at 425° to 475°C. for 15 minutes or until no
further fumes are evolved. Allow the crucible to cool,
remove it, and add 2 to 3 m!. of concentrated nitric acid.
Evaporate gently to dryness, and cautiously decompose the
residual nitrate over a small flame. 'Then ignite at a low :t:ed
heat to constant weight. The loss in weight gives th,~ weight
of pure stannic oxide present in the residue. Calculate the
percentage of tin. #
Dissolve the residue in the crucible by digestion with a few

ml. of concentrated hydrochloric acid, and add it to the filtrate
and washings from the metastannic acid. Add 7 to 8 m]. of
concentrated sulphuric acid, and evaporate the solution care-
fully (e.g., on a low temperature hot plate) until white fumes
of sulphur trioxide are freely evolved. Ccrrnplete the deter-
mination of lead as sulphate according to the proceq.ure given
for Lead in Section IV, 87 (b) (Method A). Calculate the
percentage of lead.
Note. 1. Ridsdale's .. ,Solder, No. Ua" (one of the Analysed Samples for
Students) is an almost pure lead-tin alloy, and is suitable for this analysis.
IV, 91. ANALYSIS OF A SILVER COINAGE ALLOY

Discussion. Prior to 1.920 British silver coins contained from 90
to 95 per cent of silver, the remainder b,eing copper; subsequently
the percentage of silver was reduced to 50 and this was allQyed with
nickel and copper. Many silver coinage alloys also contain a little
zinc; the procedure to be described will therefore permit of the
separation and determination of silver, copper, nickel and zinc.
Procedure. The alloy should be Clean and grease-free. If
dirty, the surfa,ce may be cleaned with a little fine emery paper
or cloth, and then washed with dry acetone. The coin or
sheet of alloy may be cut with the shears into pieces, each
weighing 0·4 to 0·5 g. .-
(a) Silver. Weigh out accurately about 0·4 g. of the alloy
into a 600 ml. beaker, which is covered with a clock glass.
Add 5 ml. 'of water, and then concentrated nitric acid slowly
(5 to 10 mI.). Warm on a water bath,or upon a low tempera-
ture hot plate until solution i.s c.omplete (the minimum
quantity of nitric acid required for dissolution should be
used). Dilute to 300 ml., heat to 70°C, and add approxi-
.mately 0'2N hydrochloric acid. drop wise and with constant
stirring until precipitation is complete. Complete the deter-
mi~ation as in Section IV, 15, 'and weigh the silver chloride.
Reserve the filtrate and washings (i).
(~) Copper. Add 3 to 4 m!. of concentrated sulphuric acid
to the combined filtrate and washings (i) from the silver
estimation, and evaporate until dense fumes of sulphur
trioxide appear. Allow the solution to cool, wash down
the sides of the vessel with 10 to 20 m!. of water, dilute to 100
mL, and almost neutralise the solution with 1 : l' ammonia
solution. Then add 5 m!. of cold, freshly-prepared, saturated
sulphurous acid solution, followed by 10 per cent ammonium
thiocyanate solution slowly and with constant stirring from a
burette until present in slight excess (ca. 10 mI.). Complete
the estimation as in Section IV, 19A, and weigh the precipitate
as cuprous thiocyanate. Keep the filtrate and washings (ii)
for the estimation of nrckel and zinc.
(c):Nickel. Evaporate the combined filtrate and washings
(ii) to 50 ml., add 35 m!. of concentrated nitric acid and 15 m!.
of concentrated hydrochloric acid, and evaporate the solution
nearly to dryness. Tnis will destroy the excess of thiocyanate ..
Rinse the clock glass, and evaporate to ch-yness on the water
bath in order to expel the ex~ess of nitric acid. A-dd 1 ml. of
qmcentrated hydrochloric aCId, 150 ml. of water, and 5 g. of
ammonium chloride (to r~tain the zinc in solution), and then
1 : 1 ammonia solution until slightly ammoniacal. Precipi-
tate the nickel as the dimethylglyoxime complex; follow the
procedure given under German silver (Section IV, 89 (e) ).
(d) Zinc. Determine the zinc as the ammonium phosphate
and the pyrophosphate exactly as described in Section IV,
89 (e).
Finally, sum up the percentages of silver, copper, nickel, and
zinc (if present).
IV, 92. ANALYSIS OF STEEL
Discussion. Plain carbon steel contains a certain amount of car-
bon, silicon, sulphur, phosphorus, and manganese. For special
purposes, varying amounts of other elements, such as chromium,
vanadium, molybdenum, tungsten, titanium, nickel,cobalt,zirconium
and copper, are added. Iron itself is rarely determined in steel;
only the subsidiary elements are ~stimated, for these are of great
'practical importance. GeneI;ally a number of methods are avail-
able for the determination of each element. The procedure em-
ployed will depend upon the object in view; the more accurate
methods are usually less rapid. In this book only a selected number
.of exact methods for determining the most important elements
which are found in steel will be described. For a more detailed
treatment of the subject, the reader is referred to the numerous
special works on iron and steel analy~is (see Appendix).
Procedure. A. Carbon. The total carbon in steel is deter-
mined by combustion at 1000° to 1300°C in an atmosphere of
pure oxygen. The gases resulting from the ~ombustion
(C0 2 , S02 and S03) are passed through reagents which absorb
everything but oxygen tlnd carbon dioxide, and the carbon
dioxide is finally absorbed in a U-tube containing a suitable
solid reagent (see Section IV, 76).*
• For full experimental details the reader is referred to any of the standard
text books on the subject, e.g., Lundell, Hofmann, a'nd Bright, Chemical
Analysis of Iron and Steel (J. Wiley, 1931) or to Sampling and Analysis of
Carbon and Alloy Steels (Rheinold Publishing Co., 1938).
B. Phosphorus. This element occurs as phosphide, and
must be oxidised to orthophosphate before it can be deter-
mined. Weigh out accurately about 3 g. of the steel (1) into
a 250 ml. conical flask, ·add 75 m!. of 6N (3 : 8) nitric acid, and
warm gently on a hot plate until the solid has dissolved.
Then add 20 ml. of 2 per cent potassium permanganate
solution, and boil for a few minutes in order to precipitate the
manganese dioxide; add more permanganate solution if no
precipitate appears. Dissolve the precipitate by adding a
freshly prepared solution of sodium sulphite dropwise;
boil for a few minutes to expel sulphur dioxide and oxides of
nitrogtm. Partially neutralise the nitric acid by the addition
of 20 to 25 m!. of 6N ammonia solution, and then introduce
75 ml. of ammonium molybdate reagent. Stopper the flask
with a rubber bung, and shake for 5 minutes. Wash off the
stopper, and allow to stand for 2 hours, or preferably over-
night. Filter off the precipitate of ammonium phosphomolyb-
date, and wash it moderately with 5 per cent ammonium nitrate
solution. Dissolve the precipitate in 20 m!. of 6N ammonia
solution containing 2 g. of citric acid, and collect the filtrate
in a 250 ml. beaker. Wash the filter paper 2 to 3 times with
1 : 20 ammonia solution, and fin".lly twice with 1 : 20 hydro-
chloric acid. Add a few drops of methyl red, and acidify the
solution with dilute hydroc~loric acid. Next add 25 ml. of
magnesia mixture (Section IV, 20B, Note 2). Complete the
estimation as in Section IV, 65-B, and weigh the precipitate as
magnesium pyrophosphate.
Note. 1. Any of the following samples are suitable for this estimation:
Ridsdale's .. Medium Carbon Steel, NCo. 2" (one of the.Analysed Samples for
Students); the Bureau of Analysed Samples .. Carbon Steel, ' P , or ' SI ' ..
(British Chemical Standards); British Standardised Steel Sample, No.3 (sup-
plied by the National Physical Laboratory, Metallurgy Department, Tedding.
ton, Middlesex, England).
B. Sulphur. Two methods for the determination of sulphur

in steels are generally employed. In the first, the sulphur is
ultimately weighed as barium sulphate, and is the more
accurate. In the second, the hydrogen sulphide evolved by
the action of hydrochloric acid, is absorbed in an ammoniacal
solation of cadmium chloride or of' zinc sulphate, and the
precipitated sulphide is titrated in 'an acid medium with
standard iodine solution in the presenc~ of starch as indicator.
:rhe volumetric rpethod tentatively assumes that the sulphur
is present in the form of sulphide, and that all the sulphur is
evolved as hydrogen sulphide upon treatment with acid.
Gravimet;ic Analysis 641
This is not always entirely the case so that the procedure is
not as accurate as the gravimetric method, but owing to its
greater rapidity it is very often used, particularly for routine
work. The accuracy of the method may be increased by
standardising the iodine sotution against a steel of known
sulphur content similar in characteristics to the sample being
analysed. For further details, consult any good text book on
Metallurgical Analysis (see Appendix, Section A, 3).
Procedure. Gravimetric method. Weigh out accurately
about 5 g. of the steel (1) into a covered 350 ml. conical flask
or beaker. Add 50 ml. of pure concentrated nitric acid.
Warm, if necessary, on the steam bath until the sample is
dissolved: the cautious addition of up to 5 m!'. of concen-
trated hydrochloric acid will facilitate solution in cases where
difficulty arises. When the sample has dissolved, add 1 g.
of A.R. sodium chloride (2), evaporate to dryness, and bake
gently on a hot platf( for about 10 minutes (3). Cool,-add 25
ml. of concentrated hydrochloric acid, and evaporate the
solution almost to dryness (i,e., until syrupy). Then add 5 m!.
of concentrated hydrochloric acid, 25 ml. of water, and 5 g. of
A.R. (sulphur free) zinc granules (4). Warm on the steam
bath until all the ferric iron has been reduced and evolution
of hydrogen ceases. Filter the solution through a small
filter, and collect the filtrate in a 250 to 400 m!. beaker.
Wash the residue with small volumes of 1 : 50 hydrochloric
acid (use about 50 m!. in all) .• Warm the filtrate and wash-
ings to 70 a C, and 'precipitate the sulphate by the sl,ow addition
of 10 m!. of hot 10 per. cent barium cfiloride solution, Allow
to stand overnight, and filter througl:t a What man No. 40
or No. 540 filter'paper .• Wash first with a solution conta,ining
10 m!. of concentrated hydrochloric acid and 1 g; of barium
chloride per litre, then with cold water to remove chlorides
(50 to 75 mI.). Ignite and weigh the precipitate as barium
sulphate according to the procedur.e of Section IV, 6. For
the best results a platinum crucible should be used; the
ignited residue should be treated with 1 drop of concentrated
sulphuric acid and 1 m!. of pure (e.g., A.R.) hydrofluoric acid,
evaporated to dryness, and ignited again to constant weight.
Notes. 1. Any of the following samples are suitable for practice in this
estimation: Ridsdale's" Medium Carbon Steel, ~o. 2" (one of the Analysed
Samples for Students); the Bur~au of Analysed Samples .. Carbon Steel,
• H'" or • R'" (British Chemical Standards); British Standardised Steel
Sample, 'No.1 or No.2 (supplied by the N.P.L.)-see Appendix, Section A,S.
2. The addition of sodium chloride will fix sulphuric acid as sodium sulphate.
y
.
3. The nitric acid must be removed, because of the serious coprecipitation
of nitrates in the subsequent precipitation of barium sulphate (see Section
IV,6).
4. Ferrous and zinc chlorides are not carried down to ;tny serious extent by
harium sulphate: ferric chloride is coprecipita.~ed if present in appreciable
concentration.
c: Silicon. The steel may be dissolved in nitric acid, dilute
sulphuric acid, or in a mixture of nitric and sulphuric acids;
the ..silica is reIJ.dered insoluble and purified with the aid of
hydrochloric acid as already described (Section IV, 70). The
best procedure is to employ 60 to 70 per cent perchloric acid
for dehydrating the silica; with this acid the silica is more
readily filtered, and is practically free from impurities since
the perchlorates of the metals are generally very soluble.
Both methods will be given.
Method 1. Weigh out accUl;ately about 5 g. of the steel (1)
into a porcelain dish or casserole, and cover it with a clock
glass. Add carefully 100 m!. of a mixture of 35 parts of
concentrated nitric acid, 15 parts ¢ concentrated sulphuric
acid"and 50 parts of wat~r. Warm until reaction ceases.
When the sample has dissolved, evaporate the solution until
dense fumes of sulphur trioxide are evolved fDr 2 to 3 minutes;
take precautions to avoid spattering. Allow to cool, add 10
m!. of concentrated hydrochloric acid, and wash off the
cover jnto the dish. Treat with 150 ml. of water at 40°C,
warm, but do not boil, to dissolve the, salts. Filter at once
through a Whatman No. 40 or No. 540 paper; use a rub~er
policeman to dislodge any pa"rticles adhering to the sides o'f
the vessel .• Wash the-precipitate with cold 1 : 20 hydro-
chloric acid until the ferric salts have l::ieen removed, and then
several times with hot water. Place the papeJ: and precipitate
in a weighed platinum crucible, heat ~lowly to char the paper
and burn off the carbon, and then apply the full heat of a
Meker type of burner. Ignite to constant weight. In accurate
work the weight of pure Si0 2 present in the residue should be
ascertained by treatment with a little concentrated sulphuric
and hydrofluoric acids (for details, see Section IV. 70).
Method 2. Weigh 'out accurately about 2'Q g. of the steel
into a 400 ml. beaker, provided with a clock glass cover. Add
40 m!. of 6N nitric acid. When the sample has dissolved,
rinse and remove the glass cover, add 25 to 30 m!. of A.R. 60
to 70 per cent perchloric acid (DANGER, see Section IV, 70),
ahd evaporate on a hot plate to copious fumes of perchlqric
acid. Replace the cover on the beaker, and heat so that the
perchloric acid condenses and runs down the sides of the
Gravimetric A nalysis 64~
beak~Jor 15 minutes. Do not allow the contents of the vessel

to becd.\ne pasty or solid, since if this occurs the separation of
silica i~ always incomplete; add more perchloric acid, .. if
necessa..,.. Allow to cool somewhat, dilute with 125 mI. of
hot wqter, and heat to boiling. Filter off the precipitated
silica, and complete the determination as in Method [.
Note. 1. The same sample as was employed for the estimation of sulphur
may be employed, except that British Standardised Steel Sample, -No. 12
(Cast Iron Standard Sample for Si, P, Mn and S) be utilised.
D. Manganese. This element is usually determined by the
bismuthate process (Section m, 60). For practice in this
estimation, the student may employ the same sample of steel
as was elnployed for Silicon.
E. Nickel. This element is nearly always determined by
precipitation with dimethylglyoxime. Full experimental
details are given in Section IV, 12B.
F. Chromium. All methods for determining chromium in
alloys <tre based upon the oxidation of the element to the
chromate, followed by the determination of the latter by
volumetric methods. Chromium is usually associated with
vanadium in steels, and a procedure for the estimation of both
elements is given under Vanadium. Two methods for
chromium will be described here; they are applicable both to
Cr-V steels and to ferro-chromium.
Method 1. Weigh out accurately about 2 g. of th_e Cr-V
steel (I), in the form of tUFnings or drillings, into a 600 m!.
beaker, and dissolve it in 60 m!. of a solution containing 45 m!.
6f water) 12 m!. of cqncentrated sulphuric acid and 3 mI. of
pure syrupy phosphoric acid (2). Heat until _ef!ervescen.ce
nearly ,"eases, then add 5 ml. of concentrated nitric acid (3),
and boil until solution is complete and oxides of nitrogen are
expelled. If any residue of tungstic acid appears, filter it off,
and wash it thoroughly with dilute nitric acid. Dilute, to
about 250 mI., and add 10 to 15 m!. of O'IN silver nitrate .and
15 m!. of a freshly-prepared 8 per cent ammonium persul-
phate solution, and boil (4). If a permanganate colour does
not develop at first f(om the manganese in the steel, add more
silver nitrate and ammonium persulphate solutions. Con-
tinue the vigorous boiling for 8 minutes~to decompose the
excess of persulphate (5). Then add 2 m!. of 1 : 1 hydro-
chloric acid, and boil for 10 minutes after the permanganate
colour, or any precipitated' manganese dioxide, has disap-
peared: any chlorine present is thus remqved. Cool, dilute
to 400 ml., add O·IN ferrous ammonium persulphate until all
the yellow colour disappears, thus indicating the complete
reduction of the chromic and vanadic acids (6), and add a
further 10 ml.: note the exact volume of ferrous solution
added. Titrate the excess of ferrous. solution with stp.ndard
O'lN potassium permanganate until a faint pink colour is
obtained which persists for 1 minute wIth constant stirring
(7). From the volume of ferrous solution used, calculate the
percentage of Cr in the sample of steel. ,
Notes.!. The Bureau of Analysed Samples" Chromium-Vanadium Steel
• V ' " is suitable for practice in this determination. If ferro-chromium is
used (see Method 2). only about 0·25 g. should be weighed out.
2. The objec't of the phosphoric acid is to decolourise the iron salts that are
formed. .
3. Evaporation with nitric acid is necessary in order to break up the black
insoluble carbides of chromium and vanadium which are left after the non-
oxidising attack with the mixed acids. •
4. PersuIphate oxidises chromium in the presence of. silver nitrate(catalyst)
(Cr.(SO.l. + 3(NH.).S.0. + 6H.0 = 2H,CrO. + 3(NH.).SO, + 6H sSO.).
any unoxidised vanadium
(2(VO)SO. + (NH.l.S,O. + 6H.0 = 2H.VO. + (NH.laSO. + '3H.SO,)
and manganese
(2MnSO. +- 5(NH,).S.0. +- BR.O = 2HMnO, +- 5(NH,hSO. +- 7 H 2SO,).
r.. 2(NH.1.S.0, +- 2H.0 --:. 2(NH.).SO. + 2H.SO. + 0,.
heat
Partial decomposition of the permanganic acid also occurs:
4HMnO, + xH.O -l- 4MnO,. y H.O + 30 •.
G. 2H,CrO. + GH,SO, + 6FeSO. = Cr,(SO.l. + 3Fe.(SO')1I + BH.O ;
2H~VO. + 3H,SO, + 2FeSO, = V,O.,(SO.). + Fe,(SO,). + 6H.0.
7. In the back ti~ration with permanganate solution, the reduced vanadium
is reoxidised :
2KMnO, + IO(VO)SO{ + 22H sO = K,SO, + 2MnSO, + 10HaVO;t 7HaSO ••
as is also thel excess of ferrous solution. It is therefore clear that the sole
function of the permanganate solution is to react with the excess of ferrous iron
and of the vanadium that has been reduced. and consequently if the perman-
ganate and ferrous solutions are of equal normality, the difference between the
volumes used represents the volume of ferrous ammonium sulphate solution
req~ire'd to reduce the chromic acid formed from the chromium present in the
sample.
Method 2. Weigh out accurately about 0·2 g. of ferro-
chromium (8) into a tall 600 ml. beaker, 'add 20 m!. of con-
centrated hydrochloric acid, cover with a clock glass, and
heat for 15 minutes. Add 20 to 25 ml. of pure 60 to· 70 per
cent perchloric acid (DANGER, see .Section IV. 70). and boil
for 30 minutes after the hydrochloric acid has been ex-
pelled. Cool somewhat, add a few m!. of water to dissolve
the chromium trioxide, and reoxidise by boiling for 15
minutes after the water is expelled. If any undissolved
particles of metal remain, pour off most of the solution, add
a few ml. of 60 to 70 per .s:ent perchloric acid to the residue,
and boil until solution is complete. Add the main decanted
liquid, evaporate off the .water, and boil for 5 minutes. Cool
somewhat, add an equal volume of water, and boil for 3
minutes to remove chlorine. Cool, dilute to 200 mI., add
excess of O·IN ferrous ammonium sulphate, and titrate back
the excess of the latter with standard O'IN pot,!-ssium
permangan~te, allowing 1 minute for the end point.
Note. S. The Bureau of Analysed.Samples .. Ferro-Chromium, No. 203"
(one of the British Chemical Standards) is slfitable for this estimation. If a
Cr-V steel is used, weigh out accurately about 2 g., and dissolve it directly in
20 to 25 mI. of 60 to 70 per cent perchloric acid. With ferro-chromium, the
chromium trioxide initially formed tends to cover the particles of ferro-
chromium and thus pI:event complete oxida~ion.
\
G. Vanadium. The determination 'of this element is most
often requ~red for chromium-vanadium steels: one method
will be -described. The steel is dissolved and treated exactly
as in F (Method 1) for the estimation of chromium in Cr-V
steel until the solution has peen titrated with standard O'IN
potassium permanganate. Then proceed as follows.
Add a moder~te excess of O·IN ferrous ammonium sulphate
(ca. 10 mI., but note exact volume) (1) to the cold solution.
Then introduce 8 ml. of a freshly-prepared 15 per _cent am-
monium persulphate solution, and stir for I minute to oxidise
the excess of ferrous salt (2). Titra'te with standard O'IN
potassium permanganate until a pink colour persists for 1
minute after stirring (3). Calculate the percentage of vana-
dium.in the, alloy.
Notes. 1. The excess of ferrous solution reduces the vanadium to the
quadrivalent state:. •
2K3VQ, + 3HaSO, + 2FeSO, = 2(VO)SO, + Fe.(SO,). + 6H 20.
No changes occur in the valencies of the other elements.
2. The only reaction which occurs in the cool solution in the absence of a
soluble silver salt is the quantitative oxidation of the excess of ferrous sulphate
solution: .
2FeSO. + (NH.12S,o. = Fe.(SO.). + (NH.).SO,.
'The net result" of the two treatments is the reduction of the vanadium to the,
quadrivalent state.
3. The reaction is :
2KMnO. + lO(VO)SO, + H 2SO, + 22H.o
= K.SO, + 2MnSO, + 10H.VO.. + SH,SO,.
No sigIiificaI1t oxidation of chromium or decomposition of persulphate occurs
during the short time allowed for the development of the end point with per-
manganate in the cold solution.
H. Molybdenum. The most satisfactory method for the
gravimetric determination of molybdenum in steel is to
64() Quantitative Inorganic A nalY5is
precipitate the molybslenum with a-benzoinoxime, !lnd ignite
the resultant complex to, and weigh it as, molybdic .oxide
MoO a (see Section IV, 23B).
Weigh out accurately about I g. of the molybdenum steel
(1) and dissolve it in 50 ml. of i : 6 sulphuric acid: treat the
solution with the minimum volume (ca. 1 li!l.) required to
decompose carbides and to oxidise the molybdenum. Boil
the solution to expel oxides of nitrogen, and filter the liquid I
if it is not perfectly clear. Dilute to 100 mI., cool, and heat

with sufficient ferrous ammonium sulphate solution to reduce
vanadic and ~hromic acids. Cool to 5° to 10°C, add 10 mi.
of th~ reagent (2 mi. of a-benzoin oxime in 100 ml. of alcohol)
with constant stirring, followed by the addition of sufficient
bromine water to impart a pale yellow colour to the solution,
and finally a few more m( of the reagent. Allow the mixture
to stand for 10 to 15 minutes with occasional stirring,. and
then filter off the molybdenum complex. Complete the
estimation as in Section IV., 23B; use 150 ml. of the special
wash- liquid. Ignite to constant .weight in an ele~tric muffle
or crucible furnace at 500° to 525°C. If tungsten is suspected,
proceed as described in the Discussion to Sectjon IV, 23B.
Note. 1. The Bureau of Analysed Samples" Alloy Steel' B ' .. (one of the
British Chemical Standards) i& suitable for practice in this estimation.
I. Tungsten. Practice in this estimation may be obtained
by the use of a chromium-tungsten steel. The tungsten
is precipitated as tungst[c adS. with the aid of hydrochl?ric
and nitric acids, but it is difficult to collect it quantita-
tively and wash it completely because of the tendency of
tungstic acid to form colloidal solutions. This difficulty
may be overcome .by the additiop. of cinchonine-hydrochloric
acid solution to the suspension (compare Section IV. .51B).
The precipitate is, however, far from pure; it cont_ains co-
precipitated silica and other slJ,bstances. The former is
removed by volatilisation with sulphuric and hydrofluoric
acids, and the amount of the latter (chiefly ferric oxide)
is determined by a double fusion with sodium carbonate, the
W0 3 forming soluble sodium tungstate.
Weigh ouf accurately about 1 g. of the steel (1) into a 400
mi. beaker provided with <: clock glass cover, add 50 mi. of
concentrated hydrochloric acid, and heat until the alloy is
decomposed. Remove the beaker from the source of heat,
and add cautioltsly 5 to 7 ml. of concentrated nitric acid.
Heat to boiling for 5 minutes in order to oxidise all the iron
and to convert the tungsten into bright yellow tungstic acid.
Gra~imetric Analysis 647
Dilute to·lOO mI., introduce .5 'm!. of einchonine solution (2),
and digest slightly below the boiling point for :~o to 4.5 minutes,
stirring occasionally. The addition of a Whatmap. accelerator
is advantageous. Filter off the precipitated tungstic acid
through a Whatman No. 40 or No. 54<Tfilter paper,.and wash
the precipitate thoroughly with the cinchonine hydrochloride
solution diluted 50-fold. Transfer the precipitate and filter
• to a platinum crucible, dry the paper, char it, and burn off
the carbon at tile lowest possible temperature. Cool. the
crucible, add 1 drop of concentrated sulphuric acid and 2 ml.
of A.R. hydrofluoric acid. Evaporate off the hydrofluoric
acid and. expel the sulphuric acid: finally dry at 7.50° to 850°C
in an electric muffle futnace to constant weight (3). The
weighed resi!iue will be free from silica, but will consist of
il11pure WO s (ferric oxide and other oxides may be present as
impurities). Fuse the weighed residue with 5 g. of A.R.
anhydrous sodium carbonate, extract the cold melt with 75
to 100 ml. of water, and filter off any. insoluble matter on a
quantitative filter paper: rinse out the crucible carefully, and
wash the paper and residue thoroughly with hot water. Place
the filter in the crucible, char the paper, and burn off the
carbon; fuse the residue with 1 ~. of A.R. sodium carbonate.
Extract the cold melt with 25 ml. of water, filter, and wash
the crucible and paper with hot water. Place the filter pap~r
in the crucible, burn, and ignite to constant weight; this will
give the weight-of ferric oxide, etc. Deduct this weight from
the weight of impure tungstic oxide. Calculate the percen-
tage of tungsten in the alloy.
Notes. 1. Ridsdale's " Alloy Steel, No. 3c" (one of the Analysed Samples
for Students) or the Bureau of Analysed Samples " Alloy Steel, • W ' .. is
'satiSfactory for practice in this determination. '
2. Prepared by dissolving 12·5 g. of cinchonine in 100 ml. of 6N hydro-
chloric acid.
3. Alternatively, heat the covered crucible tJ'redness.
J. Titanium: This element may be present in steels up to a
few tenths per cent, and is then best determined colorimetric-
ally. Essentially one method consists in adding hydrogen
peroxide to a sulphuric acid solution of the steel, and compar-
ing the resultant yello~ colour with a standard solution
containing known amounts of titanium heated in the same
way (compare Section V, 12).' • .
Ferro-titanium usually contains 10 to 25 per cent of Ti,
and a method for its gravimetric determination will be
described. -
Weigh out accurately 0·1) t6 1 g. of ferro-titanium (1), and
dissolve it in 50 ml. of 1 : 4 sulphuric acid in a 600 m). beaker.
When effervescence ceases, evaporate the solution until
fumes of sulphur trioxide appear. Cool, add 100 m!. of
wa:ter, and heat to dissolve iron salts. Filter, and wash the
precipitate thoroughly with dilute sulphuric acid; reserve
the filtrate (i). Transfer the filter and precipitate to a weighed
platinum crucible, burn off the carbon, then ignite for 10 to
15 minutes, and weigh. Add a few drops of concentrated
sulphuric acid and 2 ml. of A.R. hydrofluoric acid, evaporate
to dryness, cool, and weigh. The loss in weight is due to
Si0 2 • Reserve the residue. .
Dilute the filtrate (i) to 400 mI., add ammonia solution
until a faint precipitate is obtained, then add successively 2
ml. of 1 : 1 hydrochloric acid, H) rill. of glacial acetic acid, and
10 g. of sodium thiosulphate dissolved in 50 m!. of water.
Bbil for 10 minutes, allow to settle, filter, and wash several
times with 2 per cent acetic acid. Transfer the filter and
precipitate to the, platinum Crucible containing the residue
from the silica determination, burn off the carbon, then add
5 g. of A.R. anhydrous sodium carbonate, and fuse until a
quiet melt is obtained. Allow to cool, transfer the 'crucible to
a beaker, add 50 m!. of water, and heat until the fused mass is
disintegrated and no hard particles remain. Remove the
crucible and wash it. Filter and wash well with hot water (2).
Dissolve the residue on the paper in 25 ml. of hot 1 : 1 hydro-
chloric acid. and precipitate in the same way as before by
adding ammonia solution, acetic acid and sodium thiosul-
phate. Filter, wash with 2 per cent acetic acid until free
from chloride. Transfer again to a platinum crucible, heat
gently at first to burn off the paper and oxidise the carbon, and
finally ignite to constant weight. Weigh as Ti0 2 (3).
Notes. 1. The Bureau of Analysed Samples" Ferro-Titanium" (one of the
British Chemical Standards) is suitable for this estimation.
2. Sodium titanate is insoluble. whilst sodium phosphate and aluminate
dissolve.
3. The method of precipi,tation with tannin and antipyrin (Section IV, 40B)
is a considerable improvement upon the above procedure. The student
should employ this method also, and compare the results obtained. The best
method is probably that based upon precipitation with para-hydroxy-pbenyl-
arsonic acid (Section IV, 40D).
K. Cobalt. This element is present in certain Cr-V-W-Co

and Cr-V-W-Co-Mo steels. It may be determined with a-
nitroso-,8-naphthol or, preferably, with, a-nikro-.8-naphthol.
Both methods' will he described.
Gravimetrio Analysis 649
Method 1. Weigh out sufficient of the steel to contain 0·02
to 0·03 g. of cobalt (1), dissolve it in 20 m!. of 3N sulphuric
acid, and oxidise the iron by the careful addition of nitric
acid, avoiding a large excess. Evaporate to a syrupy con-
sistency, dilute to 200 mI., and slowly add a suspension of
zinc oxide until all the· iron is precipitated; avoid'a large
excess of zinc oxide. Cool, and filter through a dry filter
paper. Add 2 mL of concentrated hydrochloric acid, and
evaporate to about 10 ml. Transfer quantitatively with a.
little water to a 400 ml. beaker, add 10 drops of 100-volume
hydrogen peroxide,. followed by a slight excess of 2N sodium
hydroxide to precipitate the cobaltic hydroxide. Dissolve
the prec.ipitate in 70 m!. of glacial acetic acid, and dilute to
200 m!. with hot water. Precipitate the cobalt by the
addition of freshly-prepared a-nitroso-.8-naphthol solution,
and heat to boiling. Complete the determination as des-
cribed in Section IV, 33A. Weigh as Co(CloHe02N)3,2HzO.
Note. 1. For practice in this estimation, the student may employ 0·5 g.;
accurately weighed. of the Bureau of Analysed Samples .. Alloy Steel•• W ...
(a Cr-V-W-Co steel).
Method 2. Weigh out sufficient of the steel to contain 0·02
to 0·03 g. of Co., and dissolve it in a porcelain dish on the
water bath in 10 to 20 m!. of 1 : 1 hydrochloric acid. When
the re'action is- over, cover with a clock glass, add 1 to 2 m!.
of q:mcentrated nitric acid, and evaporate to a syrupy con-
sistency. Moisten the residue with a few drops of -hydro-
chloric acid, and transfer quantitatively with the aid of 200
m!. of water to a 500 to 600 m!. beaker. Heat to boiling, and
stir in zinc oxide suspension until the ferric hydroxide coagu-
lates, but avoid a large excess of zinc oxide. Allow to cool,
filter through a dry filter paper, and evaporate the filtrate
with 1 to 2 m!. of concentrated hydrochloric acid to about 10
ml. Transfer the solution with a little water to a 400 m!.
beaker, add 10 drops of 100-volume hyqrogen peroxide, and a
slight excess of 2N sodium hydroxide. Dissolve the .precipi-
tated cobaltic ~ydroxide in 70 m!. of glacial acetic acid, and
dilute to 200 ml. with hot water. Precipitate the cobalt by
the-addition of 30 ml. of a-nitro-.8-naphthol reagent, and heat
to boiling. Complete the determination as described in
S~ction lV, SSB. Weigh as Co(CloHe03N)3.
L. Zirconium. An excellent reagent for the determina-
tion of zirconium in steels is n-propylarsonic acid (see Section
lV, 410). The experimental details for a complex steel,
containing 1 to 2 per cent of Zr, will be given.
y*
Dissolve 3 g. of the steel in 40 m!. of concentrated hydro~
chloric acid and add sufficient concentrated nitric acid to
oxidise the iron. Evaporate to dryness, moisten the residue
with concentrated hydrochloric acid, e'liaporate again to
dryness, and heat for 30 minutes at 10fJ o to 110°C. Treat the
residue"with 10 m!. of concentrated, hydrochloric acid, and
warm gently' until solution is complete. Dilute to 100 mI.,
fIlter whilst hot, and wash the residue (silica, zirconium
phosphate, etc.) well with hot water. Reserve the filtrate (i),
and' treat thc residue as follows in order to recover the small
quantity of zirconium pr,esent. .
Ignite the residue in a platinum crucible, and volatilise the
silica with sulphuric and hydrofluoric acids in tHe usual
manner (Section IV, 70). Fuse the residue with sodium car-
bonate, extract with water, filter, and wash well with water.
Discard the filtrate. Ignite the residue in a platinum crucible,
and fuse with a little potassium pyrosulphate. Extract with
watcr containing a few drops of dilute sulphuric acid, filter,
and wash with hot water. Add the filtrate and washings to
the main filtrate (i). •
Dilute the combined filtrates to a volume.. of 300 mI., heat
to boiling, and precipitate the zirconium with 40 to 50 m!.
of a 5 per cent aqueous solution of n-propylarsoqic acid (1).
Digest for 30 to 60 minutes, filter whilst hot, with gentle
suction, through a quantitative filter paper supported on a
filter cone (Whatman, hardened, No. 51), and wash with hot
water. Return the filter paper and precipitate to the original
beaker, and disintegrate by warming gently with 10 ml. of
concentrated hydrochloric acid. Add 100 ml. of water, and
reprecipitate the zirconium by adding 10 ml. of the n-pro-
pylarsonic acid reagent. Digest for 30 minutes, filter as
before, and wash well with hot water. Transfer to a weighed
porcelain or Main-Smith crucible, ignite gently at 'first and
then more strongly ov,er a Mekcr type of burner to constant
weight. If tin is present, add ammonium iodide, follow the
procedure of Secti"on IV, 22A, and ignite to constant weight.
Weigh as Zr0 2 •
Note. 1. This is an approximately saturated solution. The reagent may be
purchased from the Eastman Kodak Company, Rochester, N.Y., U.S.A.
M. Copper. This element is usually present to the extent
of 0·005 to 1 per cent in certain steels. Two methods will be
described for its determination-a gravimetric and a ,volu-
metric procedure. The manner of effecting soluti~n of the
steel will depertd upon whether tungsten is present or absent.
Gravimetric method. (a) Tu~gsten absent. Transfer 1 to 10
g., accurately weighed, of the steel (according to the copper
content) to a 800 ml. beaker, add 200 ml. of dilute sulphuric
acid (I : .9) and heat gently until solution is complete. (If
much chromium or vanadium is present, add a few crystals
of potassium chlQrate and boil to dissolve carbides.) Dilute
to 500 m!. with water, heat to boiling, and saturate with
hydrogen sulphide for 30 minutes as the solution cools.
Digest for about 30 minutes, add a What man accelerator
(or one-third of an ashless tablet), filter on a quantitative
filter paper, and wash 3 or 4 times with dilute sulphuric acid
(1 : 99) which has been saturated with hydrogen sulphide.
(b) Tungsten present. Dissolve up to 10 g., accurately
weighed, of the steel in 100 ml of dilute 'hydrochloric acid
(I : 1). Add 25 ml. of dilute nitric acid (1 : 1) portionwise,
and boil gently until the separated tungstic acid becomes
bright yellow. Dilute the solution to 150 ml. with boiling
wa-ter, digest for 5 minutes, filter the solution-(which contains
practically all, of the copper) and wash with dilute hydro-
chloric acid (1: 9). Add 30 ml. of dilute sulphuric acid (I :.1)
to the filtrate, evaporate just to fumes of sulphuric acid, cool
and add 100 ml. of water. If a small residue to tungstic
-acid separates, filter ~nc). wash with dilute sulphuric acid
(1 : 20). To the cl~ar filtrate acfd 5 g. of tartaric acid, neu-
tralise with ammonia solution, adjust the acidity to 5 per
cent by volume of sulphuric acid, heat to boiling and pass a
rapid stream of hydrogen sulphide for 10 to 15 minutes.
Digest the precipitate at 50° to 60°C for 1 hour, filter and wash
the precipitate as under (a).
Transfer the paper containing the copper sulphide precipi-
ta~e, which may also contain the sulphides of molybdenum,
vanadium, arsenic, tin, etc.,. to a Main-Smith crucible, ignite
at a temperature not exceeding 550°C (1) and fuse with 2 to
4 g. of alkali pyrosulphate. Dissolve the cooled melt in 25
ml. of dilute hydrochloric acid' (I : 9), dilute to 100 ml.,
neutralise with 5 per cent.sodium hydroxide solution and add
0·3. ml. in excess. Boil for 3 minutes, digest for ca. 30 min-
utes, filter and wash the impure copper hydroxiae precipitate
5 or 6 times with a cool 0·5 per cent solj.ltion of sodium
hydroxide. Dissolve the precipitate in dilute hydrochloric
acid (1 : 3) or in dilute sulphuric acid (1 : 3), add 3 to 5 g.-
of tartaric acid, neutralise with ammonia or sodium hydroxide
solution, and acidify with dilute acetic acid. Determine the
copper with salicylaldoxime as detailed in Section IV, 190.
Note. 1. H. A. Kar (1935) gives the following method. Heat the cooled
residue in the crucible with 10.m!. of concentrated hydrochloric acid, transfer
the solution quantitatively to a 400 m!. beaker, add 2 to 3 drops of concen-
trated nitric acid, and then treat with excess of" concentrated ammonia
solution. Hl(at to boiling and_1ilter into a 600 m!. beaker; wash with hot 1
per cent ammonia solution and discard the paper. Dilute the filtrate to 250
m!., heat to boiling, and precipitate the copper by the addition dropwise of a
2 per cent alcoholic solution of a-benzoinoxime (10 to 15 mI.). The subse-
quent experimental details are similar to those given in Section IV, 19B; the
copper is best weighed as C'u(CuHllO.N) and not as CuO as Kar recommends.
Volumetric method. This method yields satisfactory results
if the percentage of copper is greater than 0·1. If the steel
contains less than 1 per cent of copper, use 5 g. of the sample;
if the copper content exceeds 1 per cent, use 2·5 g. of the steel.
WeIgh out accurately about 5 g. of the steel into a 750 m!.
conic a! flask, add 125 ml. of dilute sulpllUric acid (1 .: 9),
and heat until reaction: ceases."* Remove from the source of
heat, add ca. 15 g. of A.R. ammonium persulphate, and boil
for 15 minutes. Dilute the solution to 400 ml. with hot water,
add 70 mr. of s~dium sulphite solution (25 g. of sodium sulphite
per 100 m!.) and. then 25 ml. of potassium thiocyanate solution
(5 g. of potassium thiocyanate per 100 ml.-). Boil for 5
minutes, allow to stand for 5 minutes, and filter ..through a
quantitative filter paper. Wash the flask, paper and precipi-
tate with cold dilute sulphuric acid (2 : 98). Transfer the
paper and precipitate to a 250 to 300 luI. J"eagent bottle, add·
30 ml. of concentrated hydrochloric acid, 20 m!. of water and
5 ml. of chloroform or carbon tetrachloride. Titrate the
copper with standard 0·025M potassium iodate solution as
detailed in Section m, 126.
IV,93. ANALYSIS OF AN ALUMINIUM ALLOY
Discussion. There are numerous kinds of aluminium alloys;
thus we have: !i) Ultra-light alloys: these have a moderate altr-
minium content, a high percentage· of magnesium, and small per-
centage of copper, zinc, silicon, etc.
(ii) Light alloys: these have a high aluminium content, together
with small quantities of copper, zinc, magnesium, etc.
(iii) Aluminium-silicon alloys: these may contain up to.20 per
cent of silicon, together with small quantities of other elements.
The ~nalysis of a typical alloy of class (ii) will be described.
Procedure. W~igh out accurately about 2 g. of the alloyt
_ • For methods of dissolving the steel in the presence of various elements,
see U.S. Steel Corporattbn, Sampling and Analysis of Carbon and Alloy Steels,
1938. p. 123 (Reinhold Publ~shing Co.).
t Ridsdale's .. Aluminium Alloy, No. 20 " (one of the Analysed Samples for
Students) is suitable for this analysis. The allpy contains ca. 87 per cent of
aluminium and small quantities of copper,lead, iron, nickel, zinc, and magnesium.
and dissolve it in 30 ml. of 20 per cent sodium hydroxide
solution, preferably in a Monel beaker (or in a large" nickel"
,crucible). When the vigorous r"eaction subsides, heat until
solution is complete, dilute with 100 ml. of hot distilled water,
filter through a. sintered glass crucible, and wa~h thoroughly
with hot water. The residue (A) contains the copper, lead,
iron, nickel, magnesium, etc., whilst the filtrate and washings
(B) contain the zinc and aluminium as the zincate and
aluminate respectively.
Lead and copper. Dissolve the residue (A) in 1 : 1 nitric
acid, and determine the lead and copper as in the analysis of
German Silver (N-ickel Silver) in Section tv, 89 (b) and (c)
respectively.
Iron. P~ as in Section IV, 89 (d).
Nickel. P..roceed as in Section IV, 89 (e).
Magnesium. Acidify the filtrate from the determination of
nickel with concentrated hydrochloric acid and add 5 ml. of
the concentrated acid in excess. Then add 25 mL of ~oncen
trated nitric acid, and .evaporate on a hot plate in the fume
cupboard either to dryness or to fumes of sulphur trioxide if
much sulphuric acid is present. This evaporation removes toe
organic matter, and incidentally also the large amount of
ammonium salts present in solution. Dilute the solution to
about 200 ml., add 10 g. of ammonium acetate, then an excess
df a 2 per cent solution of oxine in 2N acetic acid,.followed bi.
concentrated ammonia solution until the liquid is strongly
alkaline; stir the mixture vigorously during the addition.
Boil the mixture for a few minutes until the precipitate is
crystalline, filter it off on a weighed Gooch, sintered glass or
porous porcelain crucible, and wash with hot water or,
better, with 1 per cent ammonia so~ution. Dry at 130°
to 140°C to constant weight, and weigh as Mg(C 9H 70N )2'
Alternatively, dissolve the washed precipitate in 2N hydro-
chloric acid, and titrate the magnesium with standard potas-
sium bromate solution as in Section m, 138, Procedure B.
Zinc. Determine the zinc electrolytically in the filtrate and
washings (B) obtained by solution of the alloy·; use the
method described in Section IV, 84.
• The metallic residue (A) may contain a little zinc. If it is desired to
recover this, use another sample of the alloy for the determination of zinc.
DissolvcHhe residue (A) in a little hot I : I nitric acid, filter, wash with hot
water, render the filtrate and washings alkaline with aqueous sodium hydrox-
ide, boil for 10 minutes, filter off the hydroxides of iron, nickel, etc., and add
the filtrate to the main solution (B) for the determination of zinc.
654 Quantitative f 11 organic A n.alysis
Alternatively, acidify the solution (B) with dilute sul-
phuric acid, and determine the zinc volumetrically by titra-
tion against standard potassium ferrocyanide solution with
diphenylbenzidine as indicator (Section m, 145).
Silicon. Dissolve 0·5 g. to 1·0 g., accurately weighed, of the
alloy in 1-15 ml. of 30 per cent sodium hydroxide solution (pre-
pared from the A.R. solid) contained in a covered Monel
metal beaker or in a large "nickel" crucible. When the
vigorous reaction subsides, evaporate slowly on a hot plate
to a volume of about 5 ml. If the solution is' dark, add
2 to 3 ml. of 10-volume hydrogen peroxide to accelerate
oxidation, and again reduce the volume to 5 ml. Neutralise
the concentrated alkaline solution cautiously with r : 1
hydrochloric acid, and transfer quantitatively to a Pyrex
beaker. Add 20 ml. of 60 to 70 per cent perchloric acid
solution (CAUTION, see Section IV, 70); evaporate to copious
fumes, dilute with 50 ml. of hot water, add a Whatman
"accelerator (or one-third. of a Whatman ashless tablet),
II
heat to the boiling point, and filter immediately:' wash the

precipitate with warm 1 per cent hydrochloric acid. Add 10
ml. of 60 to 70 per cent p~rchlorlc acid solution to the filtrate,
"'fume," and filter as before. Dry the .filter ..papers and pre-
cipitates at 100°C, transfe!: to a platinum crucible, and ignite
at ca. 1000 0 e. Add a few drops of concentrated, sulphuric
acid, and ignita t9 constant weight. Treat the residue with
'sulphuric and hydrofluoric acids as in Section'IV, 70, and
ignite again to constant weight. The loss in weight Wl'1
give the silica content of the residue. .
.
lV,94. ANALYSIS OF A LIMESTONE OR DOLOMITE
Discussion. Limestone consists essentially of calcium carbonate,
but it often contains magnesium in amounts which vary from traces
to that present in dolomite; the'latter is strictl;y the equimolecular
compound of calcium and magnesium carbonates, CaCOa,MgCO s'
It is of course impossible to differentiate sharply between the various
varieties, but it is usual to apply the name dolomite to all carboJ;late
rocks that contain sufficient magnesium carbonate to approximate
to the above theoretical composition more or less closely. The
procedure to be described is applicable to the analysis. of any of the
carbonates from CaCO a to taCOa,MgCO a.
The usual minor constituents of limestone or dolomite are iron,
aluminium, titanium, manganese, silica (either free or combined as
clay or felspar), sulphur (as pyrites FeS2 , or gypsum CaS0 4), sodium,
potassium, and· organic matter. The minerals themselves have
many industrial uses, and for judging their suitability for any
Gravimetric A llaZvsis
-_ fi55
particular pu~pose a'complete analysis, involving th.r determination
of all the substances just mentioned, is rarely necessary. A shorter
analysis will supply all the information that is required. The
shorter technical analysis includes the. following determinations, all
of which are usually made on the same sample:
A. L,Oss on ignition.
B. Impure sili~a or insoluble matter.
C. Combined oxides" R 20 3 ." This is the residue obtained upon
ignition of the precipitate produced by ammonia solution, and con-
sists chiefly of Al 2 0 a + Fe 2 0 a together with some TiO z, Mn a0 4 and
PP5 (chiefly as AlP0 4 and FeP0 4).
D. Calcium oxide.
E. Magnesium oxide.
F. Carbon dioxide (a separate sample is used for this determination) .
The experimental details to be described below will not, however,
De confined to the determination of the constituents of technical
interest; the mode of estimation of some of the other components
will also be indicated.
AttentiOIi should be directed to the metJ:lod sometimes employed
for expressing the results of a mineral analysis. It is usual to
express the amount of each metal as its oxide, almost always in its
·higher state of oxidation, and the amount of each anion as the anhy-
dride of the most common acid. Thus magnesium is given in terms
of MgO, and phosphorus as percentage of PPo' The fact that the.
percentage of the constituent is given as the oxide does not, however,
imply a determination of oxygen; rather, the constituent is deter-
mined in the customary manner, and the amount present is calcu-
lated as if it were the oxide. The method is particularly of value
with minerals to which no simple chemical formula can be.assigned.
Procedure. A. Loss on ignition. Weigh out accurately
about 1 g. of the ground, air-dried sample (1) into a weighed
covered crucible (porcelain or platinum), and dry at 1l0°C
for 1 hour. Cool and weigh. The loss in weight represents
hygroscopic moisture. Then ignite the .crucible and its
contents slowly (use an electric muffle or crucible furnace at
1000 to 1l00°C, or over a ~eker 'type burner). The sample
0
f
must be heated slowly at first to prev.ent mechanical loss

arising from the sudden evolution. of carbon dioxide. Heat
for 1 hour at the maximum temperature, cool, and weigh.
Repeat the ignition for half-hour periods until the weight of
the residue is constant to 0·2 mg. Remember calcium oxide
is hygroscopic, and observe the precautions given in Section
. IV, 10. Calculate the percentage loss upon ignition,
Note. 1. Ridsdale's" Dolomite, No. 9b" or': Limestone, Nos. 15b or 15c"
(these are Analysed Samples for Students) arc suitable for practice in this
analysis.
B. Impure silica, or insoluble matter. Transf~r the residue
after ignition (A) to a small covered porcelain dish or casserole,
add to the small residue in the crucible a few ml. of 1 : 1
hydrochloric acid, and rinse the solution into the porcelain
dish. * Add 5 ml. of concentrated hydrochloric acid and 1
m!. of concentrated nitric acid, and stir until the solid has
dissolved. Evaporate· the solution to dryness on the steam
b;lth or upon a low temperature hot plate at 110° to 120°C;
observe the precautions given in Section IV, 70A. MoisteR
the cIry mass with 5 m!. of concentrated hydrochloric acid.
warm, and allow to stand for 5 minutes to dissolve basic salts,
Add 20 m!. of water, and warm on the steam bath for 10
minutes; use a glass rod to break up any lumps present.
When all the soluble matter has dissolved, 'filter through a
Whatman No. 41 or No. 541 paper. Wash the precipitate
thoroughly with 1 : 100 hydrochloric acid, and finally once or
twice with water to remove most of the hydrochloric acid.
Evaporate the filtrate. and washings to dryness in the original
porcelai.n vessel,t heat the residue to HO° to 120°C, then take
up in hydrochloric acid and digest as before, using however
smaller volumes of hydrochloric acid and water. Filter off
the precipitate on a fresh filter, and wash as before. Reserve
• the combined filtrate and washings (i) for C. Place the two
silica precipitates in a weighed crucible, dry and ignite at the
full temperature of a Meker type burner. Allow to cool and
weigh. Calculate the percentage of impure silica or insoluble
matter present. .
Note. The dehydration is most simply effected by boiling with 12 to 15 m!.
of 60 to 70 per cent perchloric acid instead of evaporation to dryness (compare
Section IV, 70, Discussion).
C. Combined oxides. Dilute the filtrate (i) from deter-
mination B to 290 mI., add 5 g. of ammonium chloride (to
prevent precipitation of magnesium hydroxide), heat neatly to
boiling, add 2 drops of methyl red indicator, then 1 : I
ammonia solution until the solution is just alkaline (yellow).:
Boil for 1 to 3 minutes, filter the precipitate at once, and
• Alternatively, weigh out accurately about 1 g. of the powdered mineral
into a covered porcelain dish or casserole, and dissolve it in 1 : 1 hydrochloric
acid. When effervescence has ceased, evaporate to dryness all the steam
bath or lo-w temperature hot plate with stirring by means of a glass rod until
the residue is just dry. Continue as in B.
t The second evaporation is unnecessary if the silica content is small, and
results of only moderate accuracy are required.
! If manganese is present or is suspected, add about 5 m!. of bromine water
prior to the addition of ammonia solution: the manganese_will then precipitate
as hydrated manganese oxides. -
collect the filtrate in a 600 m!. beaker; wash the beaker and
precipitate -twice with lYot 2 per cent ammonium chloride
solution. Set the filtrate and washings (ii) aside. If the
precipitate is small and well washed, reprecipitation can be
omitted. If It is large, place the original beaker under the
fnnnel, dissolve the precipitate in a small volume of hot 1 : 1
hydrochloric acid, and wash the paper thoroughly with hot
water. Dilute the solution to about 75 mI., add a few drops
of methyl red indicator, reprecipitate with ammonia solution
as before, add a Whatman accelerator, arnd filter. Transfer
the precipitate completely to the filter, and wash with hot 2
per cent ammonium nitrate solution. Combine the filtrate
and washings (ii') with that of (ii). Place the paper in a
crucible, dry anti ignite the residue as detailed in Section
IV, 8. Calculate the percentage of mixed oxides" Rli03'"
D. Calcium oxide. Acidify the combined filtrates ({it) from
the ammonia pret:ipitation slightly with dilute hydrochloric
acid, and evaporate them to about 250 m!. Add 1 in!. 01
concentrated hydrochloric acid, and precipitate the calcium
by adding slowly· to the boiling solution 50 mi. of a hot 4 per
cent solution of ammonium oxalate, followed by 1 : 10
ammonia solution until the solution is neutral or slightly
alkaline (to methyl red). Test for complete precipitation by
adding a little more of the oxalate solution. Allow the pre-
cipitate to settle for an hour. Filter off the precipitate, and
wash .it 2 or 3 times with a cold, 0·1 per cent ammonium
oxalate solution_ Set aside the filtrate (iii). Place thE
original beaker under the funnel, dissolve the precipitate ·ill
25 to 50 m!. of hot 1 : 9 hydrochloric acid, and wash the papeI
with hot water;* Dilute to 100 mI., if necess,ary, add 5 m!.
of the 4 per cent ammonium oxalate solution, boil, then add
ammonia solution dropwise until the solution is just alkaline
. Allow to stand for 1 hour, filter through a quantitative filtel
paper, wash the precipitate with a 0·1 per cent ammoniuIT
oxalate solution un111 free from chloride (ca_ 50 mI.), ami
once with hot water to remove -the excess 'of oxalate.f Mi:x
* If difficulty is experienced in dissolving the precipitate directly on thl
filter. wash as much as possible of the precipitate into the original beaker. ane
dissolve the Small amount of the precipitate adhering to the paper by pourin!
hot 1 : 9 hydrochloric acid repeatedly through the filter. Then add a drop 0:
methyl red. heat until the precipitate has dissolved, adding more dilute hydro
chione acicl if necessary.,
t If the precipitate is to be weighed as CaCO a (the better method). filte
through a weighed Gooch or porous J,il0rcela,in crucible, and prOCeed as il
Section IV, 10.
the filtrate and washings (iii') with the previous one (iii),
and reserve the liquid (iv). Dry the paper and precipitate,
and ignite in a platinum cruciqle to constant weight. Weigh
as calcium oxide, following the procedure of Section IV, 10.
Calculate the percentage of CaO present. .
E. Magnesium oxide. The combined filtrates (iv) contain
excess of oxalate ion and large amounts of ammonium salts;
these, as already pointed out (Section IV, 11), cause the results
for magnesium, when' precipitated as magnesium ammonium
phosphate,:to be high. This error may be avoided by double
precipitation or by the removal of the interfering elements.
In the latter procedure any silica which may be present is
also removed. - By oxidation of the hot solution with nitric
acid in the presence of chloride ions, oxalate is oxidised to
carbon dioxide and water and the ammonium salts princi-
pally to nitrous oxide, nitrogen and water. Both methods will
be describetl.
Method 1. Acidify the filtrate (iv) from the determination
of calcium, and evaporate to 300 m!. Cool, add a few drops
of methyl red indicator, followed by 20 m!. of freshly prepared
ammonium phosphate reagent (25 g. of A.R. (NH')2HPO,
in 100 ml. of water). Then add pure concentrated ammonia
solution slowly whilst stirring the solution vigoro\1sly until
the indicator turns yellow'; continue the stirring for 5 min-
utes, keeping the solution alkaline by the addition of m'ore
ammonia solution if necessary, and finally add lO·m!' of
concentrated ammonia solution in excess. Allow the so'lution
to stand for at least 4 hours, but preferably overnight. Filter
off the precipitate and wash it 2 to 3 times with 1 : 20
ammonia solution: discard the filtrate. Place the beaker
under the funnel, and dissolve the precipitate in 25 to 50 m!.
of hot 1: 10 hydrochloric acid. Add 0·1 to 0·2 g. of
(NH4)2HPO" and precipitate the magnesium as before in a
volume of 150 m!. Allow to stand overnight, filter off the
precipitate, and wash it with cold I : ~O ammonia solution
until free from chloride. Complete the determination as
described in Section IV, 11, and weigh as magnesium pyro-
phosphate. Calculate the percentage of MgO present.
Method 2. To the'combined filtrates (iv) from the calcium
estimation, add 50 m!. of concentrated nitric acid, and
evaporate carefully to dryness on a steq.m bath or upon a
low temperature hot plate. Only a small residue (largely
magnesium salts) should remain. Add 2 to 3 ml. of concen-
trated hydrochloric acid apd 20 mI. of water, warm the
mixture for a few minutes until the solid has dissolved; if a
little silica remains, filter it through a small paper and wash
the residue thoroughly. Determine the magnesium in the'
filtrate as in Method 1; only one precipitation of magnesium
p.mmonium phosphate is necessary in this case. Calculate
the percentage of MgO p.resent.
F. Carbon dioxide. Follow the procedure given<in Section
IV, 76, using 0·5 to 0·6 g., accurately weighed, of the sample.
Calculat~ the percentage of carbon dioxide present in the
mineral.
Additional determinations. It is sometimes required to
estimate the iron content of the original mineral. As a first
approximation we may assume that the" combined oxides"
are composed of Fe 20 a + AI 2 0 a. This is not strictly true as
Ti0 2 and P20, (as metallic phosphates) mp.y be present, but
the error introduced .by making this assumption is usually
small. The Fe2 0 a 'may be determined directly as detailed
below, and the Al 2 0 a may then be obtained by difference. The
Al 2 0 a -may be written, if desired, as Al 20 a (+ Ti0 2 + P205t
etc.), if the constituents inside the parentheses ~re known to
be present. Alternatively, the' alumina content may be
determined directly with 8-hydroxyquinoline (for detai~s, see
Section IV,P5, 01; this' procedure applies in the presence of
titanium).
G•. Feme oxide.. The impure silica (B) may, and usually
does, contain a little of the mixed oxides. Treat the impure
silica, contained in a platinum crucible, with 0·5 ml. of ,con-
centrated sulphuric acid and 2 to 3 mI. of A.R. hydrofluoric
acid, evaporate the whole cautiously, and ignite (see Section
• IV, 70A); tlie loss in weight gives the amo~nt of pure silica
present. Fuse the residue with 3 to 4 g. of A.R. anhydrous
sodium carbonate, dissolve the melt cautiously in dilute
hydrochloric acid, boil, then precipitate with ammonia
solution a..s detailed in C. Filter off the precipitate, redis-
solve it in dilute hydrochloric acid, and again precipitate with
ammonia solution. Filter the latter precipit.ate through a
small filter paper, and transfer the paper and precipitate to
the original (platinum) crucible containing the "combined
oxides," and ignite in the usual manner. It is evident that
if only a minute residue is obtained after treatment with'
sulphuric and hydroflu.oric acids, the sodium carbonate fusion
is uimecessary except for the most accurate work.
660 Quant.itative I norgan'ic Analysis
Add 5 to 7 g. of pure anhydrous potassium pyrosulphate to
the platinum crucible containing the ignited mixed oxides,
and heat slightly above the melting point of the pyrosulphate ;
Keep the crucible covered. After about 20 minutes raise the
temperature slowly until eventually the bottom is at low
redness and no dark particles are visible (ca. 1 hour): there
must be no spattering. Allow the cl1lcible to cool, place it in a
250 to 400 ml. beaker, add 100 m!. of water and 3 m!. of con-
centrated sulphuric acid, and warm until all the solid has
dissolved with the exception of a few flakes of silica. Remove
the- crucible with a glass rod and wash it thorottghly with
water. Filter off the silica on a smal~ filter paper, and collect
the filtrate in a 350 m!. conical flask; wash the paper thor-
oughly with hot water. Reduce the solution it) a Jones
reductor (Section In, 51 B) or with stannous chloride (Section
m, 51A)* and determine toe resultant ferrous iron by titration
with standard O·lN potassium per:m.anganate. Calculate the
percentage of Fe2 Q a, and by subtracting the weight of tfie
latter from that of the." combined oxides," evaluate the
percentage of " AI 20 3 ."
H. Alternative procedure for the determination of calcium
and magnesium (molybdate method). The procedure given
in D and E involving double precipitation of the calcium as
oxalate, followed by double precipitation of the magnesium
as the ammonium phosphate is accurate, but tim~consuming.
The following method is rapid and is based upon the com-
parative insolubility of calcium molybdate and the solubility
of magnesium molybdate under the given experimental
conQitions.
Heat the neutral solution of ,calcium and magnesium salts
(volume reduced by evaporation to 40 to 50 m!.) to boiling.
Add a drop or two of concentrated ammonia solution, fol-
lowed by the ammoniuIp. molybdate reagent (for preparation,
see under Cadmium, Section IV, 18A) at the rate of 1 drop per
second until excess is present. If this rate of addition is
exceeded, or if a stirring rod is used, the precipitate adheres very
tenaciously to the glass and must be removed with a policeman.
(To detect the excess of the molybdate reagent, boilthe solu-
tiqn for 10 minutes, allow to settle, and test a drop of the
supernatant liquid wjth a saturated solution of pyrogallol
• If titranium is to be subsequently determined, the hydrogen sulphide
method of reduction (Section m, olE) must be used. Th'll methods recom-
mended will yield slightly high results in the presence of titanium, since the
reduced titanium will consume a little permanganate soluti?!1.
in chloroform; a brown colouration indicates the presence of
molybdate.) When excess of the reagent has been added, boil
until the supernatant liquid is perfectly clear (about 15
minutes), and allow to stand until cold (30 to 40 minutes).
Filter through a weighed Gooch crucible, containing a thin
layer of ground asbestos upon the main pad, wash 10 times
with water (100 m!. in all), dry at 130°C for 30 minutes, ignite
in a larger crucible (Fig. 2-33) and weigh as CaMo0 4 • Calculate
the percentage of CaO present.
Concentrate the filtrate and washings to about 125 ml., and
precipitate the magne:;ium as magnesium ammonium phos-
phate in the usual way. Weigh as Mg2P207' and calculate
the"percentage of MgO present. The excess of ammonium
molybdate present does not interfere.
IV, 95. ANAllYSIS OF A FELSPAR*

DisclIssion. The chief constituents that are normally sought for in
an ordinary analysis of a felspar. expr~ssed in petrological nomencla-
ture, are Si02 ; AI 20 a, Fe20a, Ti0 2 ; eaO; MgO; Na 20; K 20:
the loss upon heating to 105° to 110°C is also sometimes determined.
Methods for determining all these constituents are given below.
The mineral should be in the form of a fine powder (ca. 100 mesh).
If it is in the form of lumps, the method described in ·Section II, 15
employing a hardened steel plate, a percussion mortar and an agate
mortar successively may be used. However, for practice in analy-
tical methods, the sample is usUally supplied in a finely-divided form
suitable for immediate analysis.
Procedure. A. Solution of the sample. Weigh out accur-
ately 1 to 1·5 g. of the finely powdered mineral (1) into a
platinum crucible (30 m!. siz~), and, by means of a short thin
glass rod, mix it intimately with 6 to 7 g. of A.R. anhydrous
sodium carbonate. Remove any powder from the rod by
brushing it with a cam€!l's hair brush. Cover'the crucible,
and heat, at first very carefully over a low Bunsen flame, then
gradually over the full Bunsen flame, and finally over the full
flame of a Meker type burner until quiet fusion results and
decomppsition is complete. Take care that the melt does not
"boil over." When cold, place the crucible and lid in a
covered, deep 7-inch porcelain or platinum basin or in a
casserole, cover it with distilled water, ;:tnd leave overnight,
or warm on the water bath until the contents are well disin-
tegrated. When this is the case, add concentrated hydro-
chloric acid in small portions until an excess is present, and
• This is an example of the analysis of an insoluble silicate rock.
662 QuanUtative Inorganic Analysis
warm on the water bath until the evolution of carbon 'dioxide
ceases. Remove and rinse the cover glass, crucible and lid,
and evaporate the contents of the dish or casserole to comple~e
dryness on the water bath, crushing all lumps with a glass rod.
Note. 1. Ridsdale's "Felspar (Potash), No. 29" (one of the Analysed
Samples for Students) is suitable, but the U.S. Bureau of Standards Sample
No. 70 (potash felspar) or No. 99 (soda felspar) is to be preferred since a com-
plete analysis is supplied. -
B. Silica. Heat the residue from A for an hour at 1l0° to
120°C to dehydrate the silica, and then complete the deter-
mination as in Section ~,70B. At le;:tst tw<? evaporations
are required to completely dehydrate the silica. Treat the
ignited residue with sulphuric and hydrofluoric acids, and
determine the Si0 2 by difference. After weighing, set aside
the platinum crucible without cleaning.
C. Ferric, aluminium, and titanium oxides. Dilute the
filtrate from the precipitation of silica to 200 mI., add 5 g.
of ammonium chloride (to prevent precipitation of magnesium
hydroxide), then about 3 ml. of ,bromine water, and boil.*
Boil the solution to expel excess of bromine. Add 2 drops <;>f.
methyl red indicator to the boiling solution f6110wed by pure
1 ; 1 ammonia solution until just alkaline; boil for 1 or 2
minutes. Filter off the precipitate at once and collect the
filtrate·in a 600 m!. beake.r ; wash the beaker and precipitate
with 2 pet cent ammonium chloride solution. Set aside t,he
filtrate and washings (i). Place the original beaker under the
funnel, dissolve the precipitate in hot 1.; 1 hydroch~oric acid,
and wash the paper thoroughly with hot water. Dilute. the
filtrate to 150 mi., reprecipita.te by ammonia solution as
before, add one Whatman accelerator, and filter. Transfer
the precipitate co~pletely to the paper, and wash thoroughly
with 2 per cent. ammonium nitrate solution. Combine. the
filtrate and washir~gs (i') with those of 1i). Transfer th~
precipitate and paper to the platinum crucible containing the
residue from the impure silica after treatment with sulphuric
and hydrofluoric. acids, and ignij:e the residue as detailed in
Section IV, 8. Weigh to obtain the weight of
Fe 2 0 a + Al 2 0 a + :no 2 • The percentage of Ti0 2 is usually
very small, and, for many purposes, the residue may be
regarded as a mixture of Fe Z0 3 + AI 2 0 a. If time is limited,
the percentage of the mixed oxides alone may be returned.
• This will oxidise iron to the ferric state; it will also precipitate any man-
ganese present as hydrated manganese oxide and thus prevent contamination
in the subsequent precipitation of magnesium.
For the determination of the individual elements in the
residue, proceed as follows.
01. Aluminium oxide. Dissolve a portion of the residue,
accurately weighed, in sulphuric acid, dilute, and add excess
of 10 per cent sodium hydroxide solution. Filter and wash ..
Acidify the filtrate with hydrochloric acid, and determine the
aluminium by the addition of an acetic acid solution of oxine,
followed by 2N ammonium acetate (for details, see Section
IV,28B). Weigh as AI(CpHaONh Calculate the percentage
of Al 20 3 present.
02. F~rric oxide. Use the .potassium pyrosulphate fusion
method describ.ed in Section iV, 94G with a known weiglit of
the residue. Reduce with hydrogen sulphide if titanium is
subsequently to be estimated. Titrate the ferrous salt with
standard Q'IN potassium permanganate. Calculate the per-
centage of ferric oxide present.
03. Titanium oxide. Dissolve a known weight of the mixed
oxides in sulphuric acid, and determine the titanium colori-
metrically with hydrogen peroxide (Section V, 12). If
appreciable quantities of titanium are present, precipitate
the titanium from the acid'solution with tannin and antipyrin
(for detajls, see Section IV, 40B), or with para-hydroxy-
phenyl arsonic acid (Section IV, 40D), and weigh as Ti0 2 •
Calculate the percentage of titanium oxide present.
D. Calcium oxide. Acidify the combined filtrate. and
washings (i + i') from the ammonia precipitation with hydro-
chloric acid, evaporate to about 150 mI., heat to boiling and
precipitate the calcium by means of ammonium oxalate and
ammonia solution. Filter. Dissolve the precipitate in
dilute hydrochloric p,cid, and reprecipitate the calcium by the
additioh of ammonia solution and a little ammonium oxalate
solution to the hot solution. Filter and wash with dilute
ammonium oxalate solution. Weigh the calCium either as
CaC0 3 or as CaO. Follow the experimental procedure given
in Section IV, 94D'. Calculate the percentage of CaO present.
Reserve all the filtrates and washings (ii) ..
E. Magnesium oxide. Acidify the combined filtrates and
washings (ii) from the determination of calcium, evaporate
to 150 mI., and estimate the magnesium by double precipi-
tation with diammonium hydrogen phosphate (Sectiozt
IV, 94E) as the ammonium phosphate., Since the percentage
of magnesium is small, precipitation with 8-hydroxyquinoline
may be employed with advantage (for details, see Section
664 Quantitative I norga1tic Analysis
IV, 46B and ·Section IV, 96E.) Calculate the percentage. of
MgO present.
Note. For an alternative and more rapid method for the determination of
calcium oxide and of magnesium oxide, see Section IV. 94B.
F. Potassium and sodium oxides. The 1.
Discussion.
Lawrence Smith method (1871) will be described. The very
finely-divided sampl~ is decomposed by heating it with its·
own weight of pure ammonium chloride and eight times its
weight of pure calcium carbonate. The sintered mass is
leached with water to extract the alkalis and filtered; much
of the calcium passes into solution as the hydroxide or
chloride, and silica, aluminium, magnesium, iron, etc., remain
undissolved. The extract is treated with ammonium car-
bonate solution to precipitate calcium, and ~he ammonium
salts are removed by volatilisation. The alkali chlorides
alone remain, and are weighed. Then sodium or potassium
is determined by any of the usual methods in the mixture, and
the other alkalLis estimated by difference.
Procedure. Weigh out accurately about 0·7 g. of the very
finely-divided m.ineral into an agate mortar, add an equal
weight of A.R. ammonium chloriQ_e (weighed to the nearest
2 mg.), and grind the two substances together, using a small
agate pestle. Weigh out to the nearest decigram 5·6 g. of
A.R. calcium carbonate on to a watch glass, and add three-
fourths of this, a, little at a time, to the mortar and mix
thoroughly after each addition, When the mixing is com-
plete, lay the pestle down on the watch glass contai~ing
the remainder of the calcium carbon.ate, and transfer the'
mixture to a 30 mL platinum crucible, fitted with a cover,
with the aid of a spatula (preferably of platinum): it is
advisable ta place a little of the calcium carbonate at the
bottom of the crucible in order to prevent adhesion of the
sintered material at the end of the subsequent ignition.
Rinse out the mor:tar with the remaining calcium carbonate,
and transfer to the crucible. The whole operation should be
carried out on a sheet of glazed paper, and thus avoid the
possibility of any loss in filling the crucible. Cover the
crucible with the lid, and tap the'crucible gently to cause the
powder to settle. Insert the cruciqle in a hole of a sheet .of
asb'estos (or of " uralite "); about one-third of the crucible
should extend below the asbestos. Heat the crucible with a
very small flame until the odour of ammonia is no longer
perceptible (about 30 minutes-if fumes of ammonium
chloride are evolved, the flame should be turned down).
Then heat the crucible over a Bunsen flame for 1 hour, regu-
lating the temperature so that only the lower half of the
crucible is at a red heat. Allow to cool, rinse off the cover,
add sufficient water to cover the sintered cake, and heat for
10 minutes on the water bath, adding more water if necessary.
Wash the contents of the crucible into a dish (porcelain, silica,
but, preferably, of platinum); crush .any lumps present
to powder with the agate pestle.
Add enough water to the residue in the dish to give a
volume of 50 to 75 mI., and digest on the steam bath for 30
minutes. Filter by decantation through a 9 cm. fast filtering
paper into a 600 mI. Pyrex beaker or into a large silica or
platinum dish. Add 50 m!. of hot water, heat to boiling, rub
the insoluble mass with the pestle for some minutes, and
again decant. Repeat the extractio.n four or more ti\lles,
depending upon the alkali content of the mineral. Wash the
paper with 50 ml. of hot water or, better, with hot, saturated
pure calcium hydroxide solution. Test the residue for
'complete decomposition by warming it with dilute hydro-
chloric acid; it should dissolve completely except, possibly,
for a. few white flocks of silica: no gritty particles should
remam.
Treat the filtrate with a few mI. of pure conceutrated
ammonia solutiOn, and precipitate the calcium present by the
addition of excess of ammonium carbonate solution (3 g. of
the A.R. salt in 40 m!. of water). Heat the solution to boiling,
allow the precipitate to settle, filter, and wash the precipitate
, with hot water. Redissolve the precipitate in the minimum
volume of dilute hydrochloric acid, and reprecipitate with
ammonia and ammonium carbonate solutions: this repreci-
pitation will recover the small amounts of alkali chlorides
which may have been c~precipitated.. Filter again) and wash
with hot water. CombiQ.e the filtrates and washings, and
evaporate them to complete dryness in a porcelain or silica
basin on the steam bath. * Heat the dry residue in the
covered basin uI1til all ammonium sal,ts are expelled; take
care to avoid too strong heating. The residue will be dark-
coloured, but this does not matter. Dissolve the residue in
25 m!. of hot water, and add, first, a few drops of dilute barium
chlorid~ solution to remove traces'of sulphate, and, secondly,
a few drops of ammonium oxalate solution to precipitate the
• It is better for the subsequent removal of ammonium salts to transfer the
solution quantitatively, when it has been reduced_to a small volume, to a small
platinum dish, and to c6mplete the evaporati,on in the latter.,
666 Q1tantitative Inorganic Analysis
remaining calcium. Allow to cool, and treat with a few ml.
of &mmonium carbonate solution in order to remove the excess
of barium; allow to stand overnight.
Filter off the precipitate through a small paper, evaporate
the filtrate to dryness as before, and ignite gently in a covered
basin to volatilise the last traces of ammonium.·salts. Dis-
solve the residue in water, and filter if necessary. Add 0·5 m!.
of concentrated hydrochloric acid to the filtrate; and evapor-
ate to dryness in a small weighed platinum dish or in a large
platinum crucible. Heat the dry residue carefully until tlie
chlorides just commence to melt at the edges, then cool fop
25 mi:rrutes in a desiccator, and weigh fos rapidly as pos.$ible ;
the sodium chloride of the·residue is somewhat hygroscopic.
Repeat the ignition until constant weight is obtained. This
give~ the weight of the sodium and potassium chlorides. *
Deter~ine .the potassium in the mixed chlorides by the
perchlorate method (Section IV, 48B), and the sodium by
difference. Alternatively, estimate the sodium as sodium
magnesium uranyl acetate (Section IV. 4'iC)~- and the potas-
sium by difference. Calculate the results as percentages of
K 20 and Na2 0 respectively.
G. Moisture. Dry 1 g., accurately weighed, of the f~lspar at
105° to lloac until constant in weight. Calculate tbe per-
centage loss in weight.
IV,96. ANALYSIS OF PORTLAND CEMENT

Discussion. Portland cement is an example of an artificial silicate '
whicp. is almost completely decomposed by means of acid, thus
rendering the process of fusion with sodium carbonate unnecessary.
The chief constituents, expressed in technical nomenclature, are:
silica, lime, alumina' and ferric oxide, magnesia, with small amounts
of the oxides of the alkali metals; sulphuric anhydride and carbonic
anhydride. The following constituents are. usually determined in an
industrial aI\alysis: A. insoluble matter; B. silica; C. alumina and
ferric oxide; D. lime; E. magnesia; F. Sulphuric anhydride; G.
alkalis (by difference) ; 11. loss upon ignition. ~
It should be pointed out that the ". insoluble matter" may con-
tain metallic elements, 'and for a complete scientific analysis this
should be fused with sodium carbonate or with sodium peroxide,
and the portion of the melt insoluble in water analysed for the other.
• For work demanding the highest accuracy, the residue should be dis-
solved in 5 to 10 mI. of hot water. filtered to remove any insoluble material.
and the paper washed and ignited in the platinum crucible (\r dish. The
weight of insoluble materia\. if any. is subtracted from that of the mixed
chlorides.
Gravimetric A-nalysis 667
components. This procedure is, however, unnecessary for technical
analysis. Full details for the complete anAlysis win be evident upon
reference to Sections IV, 94 and IV, 95.
Procedure. A. Insoluble matter. Weigh out accurately
about 1 g. of the finely-divided portland cement (1) into a 250
ml. porcelain dish or casserole, add 10 ml. of water, followed
by 5 mI. of concentrated hydrochloric acid, and cover imme-
diately with a clock glass; warm until effervescence ceases.
Add 40 ml. of water, digest upon a stearn bath or upon a low
temperature hot plate until the decomposition is complete.
Filter off the precipitate, and wash with water. Transfer
the filter paper and contents to a Pyrex beaker, and digl!st at
just below the boiling point wi~h 30 ml. of 5 per cent sodium
carbonate solution for 15 minutes. Filter, wash with cold
water, then with a few drops of. 1 ~ 9 hydrochloric acid, a!}d
finally with hot water. Transfer the paper and precipitate
to a weighed crucible, bum off the paper, ignite at a red
heat, allow to coot and weigh. Calculate the percentage of
insoluble matter.
Note. 1. Ridsdale's .. Portland Cement, No. 24" (one of the Analysed
Samples for Students) is satisfactory for practice in this analysis.
B. Silica. Weigh out accurately about 0.·5 g. of the sample

into a 250 ml. porcelain dish or casserole, add 30 ml. of water,
mix with a stirring rod, and then introduce 10 ml. of con-
centrated hydrochloric acid; break up any lumps with the
stirring rod.. Cover the dish, and heat on a water bath until
the cement is decomposed. Rinse the clock glass with a
little water, and evaporate to dryness on the water bath.
The clock glass should be supported on a large glass triangle
during the e~aporation; alternatively a Fisher" speedyvap "
be!lker cover may be used. Complete the determination as in
Section IV, 95B. Calculate the percentage of silica present.
Note. In a technical analySis the sulphuric-hydrofluoric acid treatment is
not usually eII1)?loyed. The residue is heated for 1 hour at 200 0 to 250°C,
then treated with 10 ml. of concentrated hydrochloric acid, warmed gently,
- diluted with hot water, filtered, and washed until free from chlorides. The
filter is dried. and ignited for 1 hour in a muffle furnace, cooled and weighed.
The weight of insoluble matter is subtracted from the weight of the residue;
the difference is taken as that of silica. The filtrates are reserved for C.
C. "Alumina aild ferric oxide." Proceed exactly as in

Section IV, 95C. The alumina and ferric oxide (and titanium
oxide, if present) may be determined as in Section IV, 95, Cl,
C2, and C3 respectively.
668 Quantitative Inorgattic Analysis
Note. In a technical analysis, the .. alumina and ferric oxide" are deter-
mined, as is also the ferric oxide. The latter is estimated in a fresh sample of
cement as follows. About 1 g. of the cement, accurately weighed, is stirred
with 50 m!. of water, 10 m!. of concentrated hydrochloric acid added, and
heated to boiling. The ferric iron is reduced with stannous chloride solution
in the usual way (Section m, S1A), treated with orthophosphoric acid and 2
to 3 drops of diphenylamine indicator, and the ferro\ls iron titrated with
standard potassium dichromate solution (Section m, 63 and m, 66; the
other indicators given in the latter Section would be more satisfactory).
D. Lime. Follow the proced).lre of Section IV, 95D.
, E. Magnesia. Follow the procedure of Section IV, 95E.
The following method based upon the precipitation of the
magnesium as the 8-hydroxy-quinolate is rapid and accurate.
Heat the filtrate and washings from the calcium determination
to boiling, add a mocterate excess of a 2 per cent solution of
oxine in 2N acetic acid, followed by 4 ml. of concentrated
ammonia solution per 100 ml. of solution; stir the mixture
vigorously during the addition. The supernatant liquid
must .be yellow at this stage, thus indicating the presence of
excess of oxine; if not, more oxine solution must be added
until the yellow colour is obtained. Boil the mixture for
several minutes until the precipitate is crystalline, and allow
to settle. Filter the precipitate through a weighed sintered
glass or porous porcelain crucible, wash it thoroughly with hot
water (or, better, with I per cent ammonia solution), and dry
to constant weight at 130 to 140°C. Weigh as Mg(C 9 H s ON)2'
Alternatively the precipitate may be. dissolved in 2N .hydro-
chloric acid, and the magnesium determined by titration
with standard potassium bromate solution as in·Section m,
138, Procedure B.
F. Sulphuric anhydride. Two methods may be used. The
first is employed for routine analysis in industry, and takes no
account of any sulphur which may be present in the insoluble
matter, and is also subject to a number of obvious errors.
The second method gives the total sulphur content of the
cement, and is therefore to be preferred.
Method 1. Weigh out accurately about 0·5 g. of the cement
into a 250 ml. porcelain dish or casserole. Follow the pro-
cedure given under B to the point where the solution is
evaporated to dryness on the water bath. Extract the
residue thoroughly with dilute hydrochloric acid, filter, and
wash the precipitate. Determine the sulphate in the hydro-
chloric acid extract by precipitation with barium chloride
solution'as described in Section IV, 6. In order to compute
the weight of barium chloride required, you may assume
Gravimetric Analysis 66U
that the cement contains about 3 per cent 01 SOa' Express

your result as percentage of S03'
Method 2. Weigh out accurately about 0·5 g. of the cement
into an iron or nickel crucible, add 2 g. each of sodium peroxide
and of anhydrous A.R. sodium carbonate, mix thoroughly,
insert the crucible in a hole in an asbestos or " uralite " shield
(this will protect the contents from the sulphur in the flame
gases), and heat the crucible gradually at first with a Bunsen
burner, and finally over a Meker burner until the contents
of the crucible are in quiet fusion. Allow to cool and, when
cold, ,extract t~e crucible with hot water in a porcelain dish
or casserole. Remove the crucible, and wash it thoroughly,
allowing the washings to run into the dish. Cover the dish,
and acidify cautiously with hydrocbloric acid. Evaporate to
dryness on the water bath. Moisten the residue with 5 ml.
of 1 : 1 hydrochloric acid, dilute with 200 ml. ot water, and
filter; wash thoroughly. Precipitate the sulphate in the
boiling solution (300 to 350 ml.) by the addition of barium
chloride solution (Section IV, 6). Filter, ignite, cool, and
weigh. Express your result as percentage of 50 s.
G. Alkalis. Use ~e J. Lawrence Smith method of Section
IV,95F.
Note. In a tcchnic~ analysis, the alkali content is usually obtained
by difference.
H. Loss upon ignition. Weigh out accurately about 1 g.
of the cement into a platinum crucible, cover with the lid,
and support in a hole in an a:sbestos or "uralite" board.
Heat gradually, and then more strongly with a Meker type
burner until constant weight is obtained (about 30 minutes).
Allow to cool in a desiccator for 25 minutes before weighing..
Exptess your result as percentage loss upon ignition.
Note. If desired, the percentage of carbonate present may be
determined by the method of Section IV• .76; use 2 to 4, g. for the
determination.
CHAPTER V
COLORDnETRIC ANALYSm
V, 1. General discussion.-The variatioI\ of the colour of a '
system with the change in concentration of some component
forms the basis of what the chemist commonly terms colori-
metric analysis. The colour is usually due to the formation
of a coloured compound by the addition of an appropriate-
reagent or it may be inherent in the desired constituent itself.
The intensity of the colour is then compared with that
obtained by treating a known amount of the substance in the
same manner.
It must however be pointed out that colorimetric analysis
is only a special case of the more general photometric chemical
analySis. The latter may be defined as the analysis which is
based upon the measurement of the quantity of light absorbed
by a coloured solution (spectrophotometry, coforimetry). or by
a white suspension (turbidimetry, Section V, 5), or of the
amount of light scattered by a suspehsion (nephelometry,
Section V, 5). In colorimetry, natural or <l,ftificial white light
is generally used as a light source and determinations are
made with a simple instrument termed a colorimeter. In
spectrophotometry, light of a definite wave length, extending
-to the ultra-violet region of the spectrum, constitutes the
source of light and this necessitates the use of a more com-
plicated, and consequently more expensive instrument,
termed a spectrophotometer. If white light is passed through
appropriate coloured filters, then light of a definite colour
extending over a narrow range of wave lengths is obtained.
By combining a 'colorimeter with suitable light filters _a
gradation photometer is. obtained. Instruments which may be
regarded as belonging to the colorimeter or gradation photo-
meter class are in common use for ordinary analytical work.
The chief advantage of colorimetric methods is that they
provide a simple means for rapidly determining minute
quantities of substances. With the development of photo-
electric instruments (Section V, SF) it is probable ,~hat photo-
metric analysis, which is comparatively simple and. rapid,
,may ultimately compete from the point of view of accuracy
with volumetric and gravimetnc analysis for ordinary
(macro) determination of elements and radicals.
670
Colorimetric Analysis 6'Z.1
Before developing the subject further, the elementary
theory of spectroph_otometry and colorimetry will be dis-
cussed.
V, 2. Theory of spectrophotometry* and colorimetry.~
When light falls 'on a homogeneous layer of a substance it m:ty
be reflected, transmitted or absorbed. If the intensity of the
'incident light of wave length A is expressed by 10 , that of the
absorbed light by la, that of the transmitted light by I, and
that of the reflected light by I" then:
10 =? II!. + t, + I,.
For aqueous solutions I, can usually be neglected by com-
parison with the other quantities; furthermore, it is also
eliminated by'th,e use of a control, such as a comparison cell,
hence:
10 = I,. + I,.
Lamoert (1760) investigatedt the relation between 10 and I,.
Beer (1852), extended the.experiments to solutions. Spectro-
photometry and colorimetry are based upon Lambert's and
Beer's laws.
Lambert's laws. 1. The amount of monochromatic light
absorbed by a body is proportional to the intensity of the
incident light. This may be expressed in the for{l1 th,at the
ratio of the intensities of the transmitted and incident light
is constant, i.e.,
I,flo = a, or I, = a,I o (1).
The constant at gives the fraction of the incident light which
is transmitted by a layer I cm. in thickness and is usually
termed the transmission coefficient.
2. When the thickness of the layer through which the light
passes increases in aritl;tmetrical progression, the intensity of
the transmitted light decreases in geometrical progression.
Thus layers of equal thickness of the same substance absorb
the same fraction of the -incident light. According to this
law:
I, = Ioae' (1 ~),
where t repres~nts the thickness of the absorbing layer.
• Spectrophotometry proper is mainly concerned .with the following
regions of the spectrum: ultra-violet, 185-400 ml< (millimicrons); visible,
400-760 ml<; infra-red, 0'76-151< (microns). In this Chapter our attention
will be confined to the visible portion of the spectrum.
1 A = 0·1 ml< = 10-'1< = 1 X 10-8 cm.
t The first investigation on this s~bject was cllrried out by Bouguer (i7~9),
but his work appears to have been overlooked. .
If 1 is the intensity of the incident light of wave length ,\,
the diminution in inteI\sity of the light on its passage through
a small thickness dt of the medium is proportional to 1 and
dt; hence:
..
-dl = k 1 dt or - ~~ = k' dt (2),
.
where k' is a proportionality factor. Integrating (2) and
putting 10 for t = 0, we obtain:
lo~ ,
lIb = k't,.
.
or, stated in other terms, I, = Io.e"-k" (3),
where 10 is the intensity of the incident light, 1, that of the
emergent'light, t is the thickness of the absorbing layer in
ems., and k' is the absorption coefficient. This is the quantita-'
tive J!xpression of Lambert's second law; it follows from (3)
that each successive layer absorbs an equal fraction 'of the
light. By changing from natural to Briggsian logarithms we
obtain:
I, = 10 , 10-°·4343."" = 10 , 10""'" (4),
where k is the extinction coefficient of Bunsen and Roscoe
(1857). Tpe extinction coefficient is usually defined as the
reciprocal of the thickness (t ems.) required to weaken the
light to IjlOth of its incident intensity. This follows from
. (4) since:
II'
o
= 0"1 = 10-'" or kt = 1 and k = ..!.
t
.
The extinction Ii is given by E = kt = log 10 11
"
Beer's law. We have thus far considered the. light absorp- •
tion and the light transmission for monochromatic light as a
function of the thickness of the absorbing lay!;r only. In
quantitative analysis,. however; we are mainly concerned
with solutions. Beer (1852) studied the effect of the concen-
tration of the coloured substance in solution upon the light
transmission or absorption. Beer found the same relation
between transmission and concentration as Lambert had
discovered betw.een transmission and thickness of the layer
(equation (3) ) :
i.e., I, = 10 . "at (5).
where c denotes the concentration. {~hen c = 1, equation
(1) is identical with equation (5)}.
Colorimetric Analysis 673
We may write:
- dl = k/ 1 dc (6)
in which dc denotes an infinitely sma}l change of concentra-
tion. Integrating equation (6) between 10 and It. we obtain
as before:
It =. 10 . 'e-k"c (7)
and It = 10 . IO-°.4343k/c
= 10 , l()-k,· (8),
Combining (4) and (8), we have:
It = lO-kic (or log 10 lIt = ktc) (9) ;
this is the fundamental equation of colorimetry and spectro-
photometry, and is often spoken of as the Beer-Lambert law.
The value of k will clearly eJapend upon the method of express-
ing the concentration. If c is stated as molar concen,tration,
k is the molar extinction coefficient. The latter is equal to
the reciprocal value of the thickness in ems. of a 1 molar
solution (c = 1) at which
It = O:} 10, since It = 10 . 10-11: when t = } and c = 1.
Expression (7) holds for varying concentrations of the
coloured component if Beer's law is applicable. The latter
may be stated in fhe form t,hat the absorption is proportional
to the numb~r of molecules in the light path.
Application of Beer's, law. Let us consider the case of two
solufions of a coloured substance having concentrations c l
and cll . These are placed in an instrument in which the thick-
ness of the layers can be altered and measured easily, and
which also allows a comparison of the transmitted light (e.g.,
a Duboscq colorimeter, Section V, 4D). When the two layers
have the same colour intensity':
Itt' = 10 . }0-kt1c, = I tz =.1 0 , lO-k!,c. (10).
Her~ tt and t2 are the lengths of the columns of the solutions
with concentrations C l and Cll respectively when the system
is optically balanced. Rence under these conditions and
when Beer's law holds:
tlcl = t 2c2 (11).
A colorimeter can therefore be employed in a dual capacity;
(a) to investigate the validity of Beer's law by varying C 1 and
C2 and noting whether equation (11) applies, and (b) for the
determination of an unknown concentration C2 of a coloun:.d
solution by comparison with a solution of known concentra-
tion (c l ). It must be emphasis"ed thabequation (11) is valid
z
only if Beer's law is obeyed over the concentration range
employed and the instrument has no optical defects.
When a gradation photometer or spectrophotometer is
used, it is unnecessary to make comparison against solutions
of 'known concentration. With these instruments the inten-
sity of the transmitted light or, better, th,e ratio Itjl o is found
directly at a known thickness t. By changing t and c the
validity of the Lambert-Beer law, equation (9), can be tested
and the value of k may be evaluated. When the latter is
known, the concentration of an unknown solution ex can be,
calculated from the formula: •
log lollt
Cx = kt _ (12).
Attention is directed to the fact tl\at the extinction coefficient
k depends upon the wave length of the incident light, the
temperature, ahd the solvent employed. In general, it is
best to work with light 01 a wave length approximating to
that for which the solution exhibits a maximum selective
absorption; the maximum sensitivity is thus obtained.
Deviations from Beer's law. Beer's law wilt generally hold
over a wide range of concentration if the structure of the
coloured ions or of the coloured non-electrolytes in the dis-
solved state does not change with the concentration. These
include salts such as chromates, permanganates and picrates,
and also many organic dyes and coloured substances. Small
amounts of electrolytes, which do not react chemically with
the coloured components, do not usually affect the light
absorption: large amounts of electrolytes may result in a
shift of the maximum absorption and may also change the
value of the extinction coefficient. If the coloured com-
pound exists in a state of partial dissociation with two or more
components, !he state of dissotiation will change with con-
centration and Beer's law will not be obeyed (e.g., for picric
acid). The law also does not apply over any appreciable
concentration range for cololJred su!lpensions.
For solutions which do not follow Beer's law, it is best to
prepare a c~libration curve with solutions of known concen-
tration. Instrument readings, for example. the scale readings
on a Duboscq colorimeter, are plotted as ordinates against
concentrations in, say, mg. per 100 ml. or 1000 ml. as abscissae.
For the most precise work each colour formation should be
calibrated throughout the dilution likely to be met with in
the actual comparison. •
Colorimetric Anaiysis 675
V, 3. ffiassification of methods of " colour" measurement
or comparison.-The basic principle of most colorimetric
measurements consists in comparing under well-defined
conditions the colour produced by the substance in unknown
amou~t with the same colour produced by a known amount of
th,e material being determined. The quantitative comparison
of these two solutions* may in general be carried out by one or
more of six methods.
A. Standard -series method. The test solution contained in
"a Nessler tub"e (Section V, 4A) is diluted to a definite volume,
thoroughly mixed ~rid its colour compared with a series of
standards similarly prepared. The concentration of the
unknown is then, of course, equal to that of the known solution
whose colour it matches exactly. The accuracy of the
method will de.f!lend inter alia upon the concentration of the
standard series; the probable error is of the order of ± 3 per
cent, but may be. as high as ± 8 per cent.
For convenience artificIal standards, e.g., Lovibond glasses,
salt solutions such as ferric chloride in aqueous hydrochloric
acid (yellow), aqueous cobalt chloride (pink), aqueous copper
sulphate (blue) and aqueous potassium dichromate (orange),
are sometimes used. It is essential to standardise the arti-
ficial standards against known amounts of -the ,substance
being determined, the latter always being treated under
exactly similar conditions. The disadvantage of this
method is that the spectral absorption curves of the fest
solutions and onhe sub-standard glasses or solutions may be
far from identical; the error due to this cause is greatly
magnified in the case of obseryer.s suffering from partial
colour blindness.
B. Duplication method. A standard solution of the com-
ponent under determination is added to the reagent until the
colour produced matches that of' the unknown sample in
the same volume of solution. This method is less accurate
than A.
C. Dilution-method. The sample and standard solution are
contained in glass tubes of the same diameter and are observed
horizontally through the tubes, The more concentrated
solution is diluted until the colours are identical in intensity
when observed horizontally through tne same thickness of
• As already pointed out (Section V, 2) it is not essential to prepare a series
of standards with the gradation photometer or the spectrophotometer, since
the concentration of the colomed component in the unknown can be computed
from the extinction coefficient, the latter being determined experimentally,
676 Quantitat;ve Inorganic Analysis
solution. The relative concentrations of the original solutions
are then proportional to the heights of the matched solutions
in the tubes. This is the least accurate method of all and will
not be discussed further.
D. Balancing method. This method forms the basis of all
colorimeters of the plunger type, e.g., in the Duboseq colori-
meter. The comparison is J;llade in two tubes and the height
of .the liquid in one tube is adjusted so that when both tubes
are observed vertically the colour intensities irr the tubes are
equal. The concentration in one of the tubes being known,.
that in the other may be calculated f~om the respective
lengths of the two columns of liquid and the relation:
C1 tl = C2 t2 (Section V. 2).
It must be emphasised again that this simple proportionality

holds only if Beer's law is applicable, and tliat the relation
holds with greater exactness if a beam of monochromatic light
(obtained with the aid of a suitabl~ colour filter) rather than
white light is employed. As a general rule it is preferable
that the solutions under comparison should n,ot differ greatly'
in concentration, and for the most accurate work an empiric-
ally constructed calibration curve should be used. A~ usually
employed with white light, the accuracy obtainable with a
Duboscq colorimeter is of the order of ± 7 per cent; the
accuracy is increased appreciably if monochromatic light
(produced with colour filters) is employed.
E. Gradation' pr step photometer method. This method is
essentially a compromise between a spectrophotometric
procedure (which emploxs monochromatic light or light of a
very narrow range of wave lengths._ca. 5OA) and the colori-
meter method (which normally utilises white light). A
spectr,ophotometer is of necessity a very expensive instru-
ment and is also somewhat tedious for analytical usage.
Colorimetric methods possess only limited acc~racy. By
employing a limited range of w?-ve lengths (ca. 250A), secured
with suitable light filters, it is pqssible to construct an instru-
ment which retains the accuracy' of the spectrophotometer
and the simplicity of manipulation of a colorimeter. Such
an in~trument is termed a gradation or S'tep photometer and
is best exemplified by the Zeiss Pulfrich photometer. With
this instrument both the percentage of the incident light
intensity which has been allowed to pass through the liquid
(i.e., IdIo) and the extinction E (Section V. 2) of the solution
are directly measured. The extinction coefficient k can then
b~ obtained by simply dividing E by the stratum thickness
(in cms.). For the measurement of concentration, all that is
necessary is to determine for a sol].ltion of known concentra-
tion C 1 the extinction coefficient kl for a. suitable light filter.
For all subsequent tests with the same filter, if C2 is the cop.-
cent ration of the unknown solution g.nd kz the corresponding
extinction coefficient, C2 is given by the equation:
C = C X -,
kl
2 1
k2
which is based upon the validity of tl].e Beer~Lambert law.
:fhe latter may be expressed in the form that the concen-
trations of two solutions are proPQrtional to their extinction
coefficients (compare equation (9) in Section V, ~). An alter-
native method of use, which is applicable also when the Beer-
Lambert law is not valid, is to carry out a series of deter-
minations in suitable graded solutions of known concentration.
and to plot a permanent calipration curve; the filter employed
(i.e., the spectral region of the measurements) and the length
of the cell must of course be noted. The procedure is rapid
and possesses the great advantage that it is unnecessary.to
prepare fresh standard solutions for each determination.
F. Photo-electric photometer method. In this method the
human eye is replaced by a suitable photo-electric cell; the
latter is also employed t(') afford a direct measure of the light
intensity and hence of the abso~ption. Instruments incor-
porating photo-electric cells measure the light absorption
and not the colour of the substance: for this reason the term
" photo-electric colorimeters " is a misnomer; better names
are photo-electric comparators, photometers or, best, absorp-
tiometers.
Essentially most such instruments consist of a light source,
a glas~ cell for the solution, a photo-electy;ic cell to r~ceive the
radiation transmitted by tpe solution and a measuring device
to determine the response of the cell. The comparator is first
calibrated'in terms of a series of solutions of known concen-
tration, and the results plotted in the form of a curve connect:
ing concentration and readings of the measuring device
employed. The concentration of the unknown solution is
then determined by noting the response of the cell and
referring to the calibration CUfYC. -
These instrnments have been made in a Ilm&er of different
form~ incorporating one or two photo-cells. With the one-cell
type, the absorption of light by the solut-iorr is usually
measured directly by determining the current output of the
photo-electric cell in relation tQ the value obtained with the
pure solvent. It is of the utmost importance to use a light
source of constant intensity-a by-no-means simple problem
-and if the cell exhibits a " fatigue effect" it is necessary to
allow the photo-cell to. attain its equilibrium current after
each change of light intensity. The two-cell type of instru-
ment is usually regarded as the more reliable in that any
fluctuation of the light source will affect both cells alike if they
are matched for their spectral response. Here the two photo-
cells illuminated by the same source of light are balanced
against each other through a galvanometer'; the test solutioil
is placed before one cell and the pure solvent before the other,
and the cur:rent output difference measured directly.
Numerous photo-electric comparators are now lJlarketed.*
Qne of the best is the Hilger "Spekker" photo-electric
absorptiometer. The latter embodies many novel fe~tures,
and, like the Pulfrich photometer, makes use of coloured
filters for definite r~nges of wave lengths (see Section V, 4F).
In the following Section it is proposed t-() discuss the most
important of the above methods in somewhat greater detail.
For a more complete treatment the reader is referred to the
special treatises on the subject (see Appendix, Section A, 3).
V, 4. A. Standard series method.. In this method colour-
less glass tube.s of uniform cross section
and with flat bottoms are usually em-
ployed. These are termed Nessler tubes.
The best variety has a polished fiat
bottom.t They are made in either the
" low" form with a height of 175 to 200
mm. and a diameter of 25 to 32 mm.
(Fig. 5-1) or as a " high" form witp. a
height of 300 to 375 mm. -and ,a
diameter of 21 to 24 mm. The solution
of the substance being determined is
FIG. 5-1. made' up to a definite volume, and
• For example: .. Photoelectric Turbidimeter and Colorimeter" (Baird and
Tatlock (London) Ltd., London, E.C.2, England); .. Electrophotometer"
(Fisher Scientific Co., Pittsburgh, Pa., U.S.A.); .. Photelometer" (Central
Scientific Co., Chicago, U.S.A.); .. Photoelectric Colorimeter and Turbidi-
meter" (Eimer and Amend. New York, U.S.A.).
t The .. doubletplane" Nessler tubes Supplied by the Fisher Scientific Cq.
of Pittsburgh, U.S.A. have bottoms which have been ground and polished
optically plane on both sides and then fused to the tubes. They are very
satisfactory in use, since they produce a uniform field ,of colour intensity.
the colot'lr is compared w~th that of a series of standards
prepared in the, same way from 'known amounts of the
component being determined. Fifty or 100 ml. of the
unknown and standard solutions are placed in Nessler
tubes, and the solutions are viewed vertically through the
length of the columns of the liquid. The concentration
of the unknown is equal to that of the standard having the
same colour. * As a general rule it will be found that the
colour intensity qf the unknowR lies between two successive
standards. Another series of standards may then be pre-
pared covering the latter range over smaller concentration
intervals. Thus, for example, in the determination of a:
particular constituent the first series of standards might
cover the range 0·1, 0·2, 0·4, 0'6, 0·8 and 1·0 mg. p.er litre,
and it is found that the colour of the unknown lies between
0·4 and 0·6 mg. per litre. The second ser·ies of standards may
then be prepared containing 0'40, 0'45, 0·50, 0·55 and 0·60
mg. per litre. Further comparison may then show that the
value lies between 0·45 and 0·50 mg. per litre, arid for many
purposes this should be returned as 0·48 mg. per litre.' If a
more accurate value is required, and provided the colour
intensity of the solution and also the apparatus employed will
permit of finer comparison, another series of standards cover-
ing the range of, say, 0·45, 0·475 and 0·50 mg. per litre may be
made up and the unknown compared with these standards.
FIG. 5-2.
* It is advisable wherever possible to make a preliminary determination of

the strength of the unknown solution by adding from a burette a solution of
the component in known concentration to a Nessler tube containing the
reagents diluted with a suitable amount of water until the depth of colour
obtained is practically th~ same as that of an equal volume of the unknown
solution also coiit'ained in a Nessler cylinder and standing at its side. A series
of standards on either side of this concentration is then prepared.
For the comparison of colours in Nessler tubes, the simplest
apparatus consists of a modified test-tuhe rack (Fig. 5-2). It
is constructed of wood, finished duB black, and is provided
(a)
FIG. 5-3;
U"k.nowo
Tube
No/aer·
(I) )
l"IG. !'i-S.
C o!.orimetric Analysis 681
This diagram showl how the light.

beams enter the bottoms of the
d-oub'e-plane Nesster tubes, pass
thl"ough both the liqu,fd bern. ex-
1In'1lned .and the standard. then
emergo and are obserwed In juata-
poslt'on Inthe eyepiec.e of the Fish ..

er Nelslerlmete!". This adjacency
greatly '.-cHltat... accurate and
r.pid c,o:)lor comparison.
{c)
FIG. 5-3.
with an inclined opal-glass refiectoL The Nessler tubes

rest on a narrow edge and do not come into contact with the
reflector. A more satisfactory method is to employ the
Fisher "Nesslerimeter." The apparatus itself is shown in
Fig. 5-:~ (a) and the optical arrangements (c) in Fig. 5-3 (b). *
It consists of a substantial cast base with a housing containing
a 15-watt daylight blue lamp for illuminating with equal
intensity both tubes being compared. There is a rotating
support, accurately mounted, which has holders for 10
Nessler tubes to contain the graded standard solutions.
The support for the Nessler tube containing the unknown
sample swings in and out of position so that it can be easily
inserted or removed. The light emerging from the two
Nessler tubes containing the standard solution and the
unknown passes into a light-tight housing at the top of the
instrument where the two beams encounter a mirror and lens
system which brjngs them side by side in the eyepiece (see
Fig. 5-3 (c)). A special filter holder is also provided for the
insertion of colour (e.g., Wratten or Corning or lena) filters.
Comp,!-rison can thus be rapidly made of the colour of the un-
known solution with that of anyone of 10 standards,
previously prepared.
There are a number of apparatus which employ permanent
glass colour standards; these are usually prepared from
* 'these diagrams are repruduced by tll(~ courtesy of the Fisher Scientific
Co ., of Pittsburgh, Pa., U .S.A.
z*
Lovibond or similar type glasses. An inexpensive model is
shown in Fig. 5-4 (Plate II). This is the B.D.H. Lovibond
nesslerimeter. * It consists essentially of a bakelite case for
holding vertically two Nessler glasses between a reflector and a
detachable rotating disc having 9 apertures containing a series
of graded permanent glass colour standards. Each disc con-
tains a series of standards designed for one particular test con-
ducted under specified conditions. Discs are available for many
of the common determinations by colorimetric methods, and
the manufacturers are continually extending the range. The
discs at present available include: ammonia (with Nessler's
reagent); 'bismuth (with potassium iodide); copper (with
dithio-oxamide); iron (with thioglycollic acid); lead (with
sodium sulphide); chlorine (with o-tolidine); nitrate (with
2 : 4-xylen-l-01) ; silicate (with ammonium molybdate) ; pH
5·2 to 6·8 (with bromo-cresol purple); pH .6·0 to 7·6 (with
bromo-thymol blue); PH 6'8 to 8·4 (with phenol red);
pH 7·2 to 8·8 (with cresol red) ; pH 8·0 to 9·6 (with thymol
blue); and pH 9·0 to ll'O (with B.D.H. 90ll ' indicator).
t
The instrument can of course be employed as a simple com-

parativ.e colorimeter as described above; the colour disc is
then removed.
B. Duplication method. This method finds its chief
application in the so-called colorimetric titration. A known
volume, say 50 or 100 mI., of the solution is placed in a Nessler
cylinder (Fig. 5-1) and a measured volume of the reagent or
reagents is then added. An equal volume of water (50 or 100
m!.) together with the same volume of the reagent is intro-
duced into another similar Nessler cylinder. For mixing the
solutions both cylinders are provided either with a glass tube
on which a flattened bulb (ca. I cm. diameter) is blown or
with a stirring rod of which the lower end is flattened to a
width of 1 cm. and over a length of several ems. The tubes
should also be provided with black or brown paper cylinders
to exclude light from the sides. The colour intensities are
compared by holding the tubes close together over a white
surface, such as sheet of opal glass or, better, in a Nessler
tube stand (Fig. 5-2). A solution containing a known con-
centration of the constituent being determined is added to the
blank solution from a burette (preferably of the micro type)
• The apparatus is manufactured by The Tintometer Ltd., Milford, Salis·
bury. in association with The Briti,;h Drug Houses Ltd., Graham Street,
London, N .1, England; it is reproduced by courtesy of the furmer company.
A similar apparatus is marketed by F. Hellige and Co.
PLATE II
FIG. 5-4
until the colours of the two solutions viewed-by looking down
into the tubes match. As a rule, if the volume of the standard
solution required to match the colour of the unknown is less
than about 2 per ceIit of the total volume, the volume change
due to the addition of the reagent may be neglected. It
may, however, be allowed for by a simple calculation, or the
estimation may be repeated by taking 100 - x (or 50 - x) ml.
of water, where x is the volume of the standard solution em-
ployed in the first titration. Several determinations should
be carried out and the positions of the tubes should be inter-
changed-thus that on the right-hand side should be put to
the left of the'observer and vice versa.
It must be emphasised that this method can only be applied
when j:He colour is independent of the mode of mixing, for in
one tube a very dilute solution of the substance to be deter-
mined is mixed with the reagent, whilst in the other tube a
comparatively concentrated solution of the substance is
mi\ed with a dilute solution of the reagent. The develop-
ment of colour should be practically instantaneous and
remain permanent during the time required for the measure-
ments; foreign substances present in the unknown should
not affect the colour. The method is, at best, only an approxi-
mate' one, but has the advantage that only the simplest
apparatus is required.
D. Balancing method. Plunger - type colorimeters. The
plunger-type of colorimeter with the two halves of the field
of view illuminated by the light passing through the unknown
and standard solutions respectively was invented by J.
Duboscq of Paris in 1854;
various improved modi-
fications of the instru-
ment have been sub-
sequently developed by
manufacturers of optical
apparatus. One of the
best of these is the Bausch
and Lomb model. Before
describing the latter,
reference must be made
to Hehner cylinders (Fig.
5-5). Thes.e are utilised
in pairs, and are the simp-
lest form of apparatus
employed in matching colours by the balancing method.
Each cylinder has a glass stop-cock about 2·5 cm. from the
bottom through which liquid may be drawn off until the
colour in the two cylinders is the same in intensity when
viewed vertically. The cylinders are graduated at 1 m!.
intervals and usually have a capacity of 100 m!.: they should
have flat, carefully ground and polished bottoms of clear glass
and be uniform in bore. It is advisable to place them in a
box so arranged that the light is reflected froin the bottom of
the latter up .through the tubes.
A Duboscq colorimeter (Bausch and Lomb, 40 mm. " bio-
logical" model*) is shown in Fig. 5-6 (Plate III) and a sectional
~ diagram in Fig. 5~7 (Plate
III). The essential principles
of the instrument will be
evident from Fig. 5-8. Light
from an even source of illum~
ination (e.g., from the special
mirror A~ adjustable on a
horizontal axis) is passed
through the two sides of the
instrument. There are inter-
posed in these two light
paths the solutions to be
tested contained in two
movable cups Band B';
_____L _.;,_ - some of the light in passing
FIG. 5-8.
through the liquids is ab-
sorbed, the amount of absorption depending upon the depth
and the concentration of the solution. The two beams of
light are now brought to a common axis by means of the
rhombohedral prisms E. Light from one cup illuminates one
half of a circular field and light from the other cup illuminates
the other half. The observing microscope F at the upper end
of the instrument, by which the observer sees both fields with
one eye, is focussed on the line of separation of the two fields.
When the eye piece is properly focussed, the observer sees the
field of view divided into two semi-circular half-fields separ-
ated by a fine line and can thus be compared with ease and
accuracy. It is important that the illumination should be
identical on both sides; the mirror surface A must be per-
• The block of Fig. 5-6, and also details concerning the use of the instrument,
are reproduced by courtesy of the Bausch and Lomb Optical Co. of Africa
House, Kings:way, London. W.C.2, England. and of Rochester, N.Y., U .S,A,
PLATE III
FIG. 5-6
c'_
FIG. 5-7
Colol'£metl'ic Analysis 685
fectly clean. It is now possible to alter the depths of the two
columns of liquid until the two halves of the field are identical
in intensity. When this condition holds and Beer's law is
applicable, the concentrations of the two solutions are
inversely proportional to their depths, which are read on the
scales of the instrument.
The all-glass cups are made of optical glass in two pieces;
the discs which form the bottoms have ground and polished
surfaces and are fused to the cylinders. The cups, which
serve to receive the solutions, are mounted movably on slides;
the cups may be raised or lowered by turning the milled heads
C and C', each of which actuates a rack and pinion. Two
solid glass plungers D and D' of optical glass, matched for
colour and having optically plane ends, are attached to the
frame of the instrument by means of metal adaptors in which
they are firmly cemented; they are located in a fixed position
on the axis of the instrument which passes through the
centres of the cups. Stops are provided which prevent the
cups and plungers being brought violently together. The
scales and verniers (not shown in Fig. 5-7 (Plate III), but one is
visible in Fig. 5-6 (Plate III) can be read to 0·1 mm.; they
measure the distance between the bottom of the cup and the
plunger.
Use of Duboscq colorimeter. The following details apply to

the Bausch and Lomb model, but can be adapted, with only
slight modification, to any make of Duboscq colorimeter.
Before the liquids are introduced into the cups, the instru-
ment must be adjusted for even illumination. If sky-light
is to be used, an unobstructed north window should be chosen
and the colorimeter placed directly in front of it. While
looking into the eyepiece, turn the mirror so that the maxi-
mum illumination is obtained. The best way to do this, for
daylight, is as follows. Prepare a dilute solution of a
coloured inorgahic salt, such as potassium permanganate.
Place equal volumes of the solution in each cup, using suffi-
cient to about half-fill each cup. Set one of the cups at .15
to 20 mm. depth and take a series of readings with the other
cup. If the average is not the same as the setting for the
first cup (within 0·1 to 0·2 mm.), change the position of the
colorimeter with reference to the illumination and repeat
until a position is found when the right and left readings are
the same. The position of the mirror and colorimeter must
not be changed from this point onwards.
686 Quantitat-ive Inorganic Analysis
The most satisfactory results are obtained, however, not
with daylight which may vary both in intensity and in colour
quality and may also be influenced by moving large objects
in the vicinity of the colorimeter, but with a lamp designed
for colorimetric work. The adjustment for illumination is
similar to that given above, except that the lamp or mirror is
adjusted instead of shifting the colorimeter. There is also a
special Bausch and Lomb colorimeter lamp available, pro-
vided with filters, etc., which requires no adjustment.
Make sure that the readings are zero when the plungers
touch the bottom of the cups. Place the standard solution in
one cup and an equal volume of the unknown solution in the
other. Set the standard at a scale reading of l5 to 20 mm.
Then adjust the unknown until the fields are mat _hed. Take
the readings on both scales. Carry out 6 to 10 adjustments
with the cup containing the unknown solution, and calculate
the mean value. If t I and t z are the average readings for the
cups containing the solutions of known and unknown concen-
tration respectively, and C 1 and C 2 are the corresponding
concentrations, then if Beer's law holds:
or
Frequently, owing to optical and mechanical imperfections

of many colorimeters, the same reading cannot be obtained
in the adjustment for illumination when the cups are filled
with the same solution and balanced. In such a case one of
the cups (say, the left one) is filled with a reference solution
(which may be a solution containing the component to be
determined) of the same colour and approximately the same
intensity as the unknown and the plunger set at some con-
venient point (about the middle) of the scale. Fill the other
cup with a solution having a colour corresponding to a known
concentration of the component to be determined, and adjust
this cup to colour balance. Take the reading and repeat the
adjustment, say, 10 times in such a way that the balancing
point is approached [) times from the lower and 5 times from
the higher side. Calculate the average reading (t1)' Remove
the cup, rinse it thoroughly and fill it with the unknown
solution. Repeat the balancing exactly as for the standard
solution, and find the average of, say, 10 readings (t2)'
If C L is the concentration in the standard solution, then
the concentratIon of the unknown solution is gIven by:
tl
Cz = CI-'
t2
(This method is comparable in many respects to the method of
weighing by substitution-see Section II, 6C.) If Beer's law
is not valid for the solution, it is best to arrange matters so
that the colour intensity of the standard lies closely to that
of the unknown.
Suggestions as to general maintenance of the colorimeter. *
1. Keep all cups and plungers scrupulously clean.
2. Before using the colorimeter be sure that it is carefully adjusted
for even Illumination, and do not subsequently change the position
of the instrument while in use.
3. Be sure that the verniers are adjusted to zero before using.
4. Wash the mirror or milk glass reflector occasionally so that it
may reflect light evenly.
5. Rinse off the plungers before putting away the instrument to
avoid contamination of other solutions.
6. Do not fill the cups to such a depth that they will overflow
when the plunger is brought into contact with the cup bottom.
E. Gradation or step photometer method. The Zeiss Pul-

frich photometert mounted on an optical bench is shown In
Eyepiece Measuring drum Cell holder Photometer lamp
FIG. 5-9.
Fig. 5-10, while FIg. 5-9 gives a diagrammatic section through

the apparatus. The optical system consists of two parallel
telescopes having a common eyepiece, and there is an arrange-
ment by means of which a number of different colour filters
* Reproduced, with permission, from" Directions for Use. " issued with the
Bausch and Lomb colorimeters.
t Figs. 5-9 and 5-10 and some of the descriptive matter are reproduced by
the courtesy of Carl Zeiss (London) Ltd., 37-41 Mortimer Street, London,
W.l, England.
688 Quantitalive Inorgan£c Al1aZysis
mq_y be introduced into the path of the light. The field of
view is divided by a vertical separating line into two semi-
circular portions, corresponding to the photometer apertures.
FIG. 5-10.
The measuring device consists of two micrometer drumheads

mounted on the openings of the telescopes (photometer
diaphragms); by turning the drums the intensity of each
half of the field can be varied by measurable amounts. The
drums are graduated so as to permit approximately the same
percentage accuracy for intensely and feebly absorbent fluids.
The concentration of a coloured solution and the absorption
of monochromatic light are connected by the Lambert-Beer
law (Section V, 2), thus permitting the determination of the
concentration from the light absorption. Nine selective
(" 5 JJ) light filters are employed to cover limited portions of
the visible spectrum. Illumination is provided by means of a
special photometer lamp with an L-bulb which yields the very
intense illumination required for use with the " 5 " filters.
For truly monochromatic light a "Hagephot" mercury
vapour lamp is available; this is provided with 3 filters for
the yellow, green and blue mercury lines.
Use of Zeiss Pulfrich photometer. A cell of suitable stratum
length (10 to 250 mm.) is filled with the unknown solution and
placed in one of the cell holders. say, the left. An identical
PLATE IV
Adam Hilge>, Ltd., London
FIG. 5-12
Colorimetric Analysis " 689
cell is then filled with distilled water, and placed in the other
cell holder. The light-filter holder is then rotated, thereby
introducing various filters successively into the path of the
light, until the colour difference vanishes. Thus with a
yellow solution, a blue filter is introduced and the field of view
appears blue on both sides but different in intensity. By
turning the drum on the right, the intensity on the left side
of the field is varied until both halves of the field of view
appear equally bright; the reading on the drum will chen
give directly the diminution which the coloured light has
experienced in passing through the unknown liquid. The
reading"on the outer (black) scale of the drum gives the inten-
sity of the transmitted light expressed as percentage of the
incident light (i.e., 1'/[0), and the reading on the inner (red)
scale gives the extinction E of the test solution. If E (= kt)
is divided by the stratum thickness (in ems.) the extinction
coefficient k is obtained. If C 1 'and kl are the concentration
and extinction coefficient respectively for a solution of known
concentration, C2 and k2 the corresponding values for an
unknown solution, then, provided the Beer-Lambert law is
valid,
k2
C2 = C
lkl ·
It is usually convenient to prepare a permanent calibration
• curve (see Section V, 3E).
F. Photo..electric photometer method. This method is well
exemplified by the "Spekker" photo-electric absorptiometer.
The optical arrangements are shown in Fig. 5-11 (a), and the
photo-electric circuit in Fig. 5-11 (b). Fig. 5-12 (Plate IV) is a
photograph of the instrument. A IOO-watt projection lamp
A, mounted in a cylindrical lamphouse and run from the
electric mains supply, is the source of light. A lens B mounted
in the housing to the right of the lamp forms a parallel beam of
light; the beam passes on through the cell containing the
solution and then falls on another lens C which forms an
image of the lamp filament on a photo-electric cell D (termed
the "indicating cell "). A calibrated variable aperture E
is mounted immediately in front of "the lens system, and
enables the intensity of the light falling on the photo-Gell to be
varied by known amounts. Since there is an image of the
filaments on tl).e cell, there is no change in the photo-cell
area illuminated when the aperture alters: only the quantity
(If light reaching the cell is controlled by the variable aperture
The scale associated with the aperture is so calibrated that,
if R is the reading corresponding to a degree of opening sllch
that the amount of light transmitted is l/a of that admitted
I 11
~I/1 I) ~ ~ [
A B
F E ADAM HILGER, l.td .•

LODdon
(a)
(b)
FIG. 5-11.
when the aperture is fully open, then R = log a. This
function (known as density) was chosen because it is approxi-
mately linear with the concentration of a solution, over small
ranges.
Light -from the lamp also falls on the photo-cell F, t~rmed
the "compensating c.ell," mounted in the housing to the
left of t}:te lamp. The amount of light falling on this photo-
cell (whith is protected by a 1 cm. water cell) can be varied
by means of an iris diaphragm G mounted in front of it.
The two photo-cells are connected in opposition across a
galvanometer so that, when the photo-electric currents given
by the cells are equal, the galvanometer shows zero deflection.
Use of "Spekker" pho'to-electric absorptiometer.* This is
l;lest illustrated by describing the procedure for making a
determination. Let us suppose that it is desired to compare
the depth of colour or, more precisely stated, the amount of
light absorbed by two liquids S1 and S2' the latter being the
more deeply coloured, i.e., the more absorbing.
• The diagrams and descriptive matter are reproduced by courtesy of the
fole manufacturers: Messrs. Adam Hilger Ltd., ~ Kings Road, London,
N.W.I. England.
Colorimetrt"c Analysis 691
1. Place S2. contained in the special cell, into the beam, and
open the variable aperture to its full extent by setting the
drum at zero.
2. Adjust the fxis diaphragm in front ~f the compensation
cell until the galvanometer shows zerO' deflection.
3. Substitute s 1 for S2 when the galvanometer will be seen
to be deflected. •
4. Adjust the calibrated variable aperture by means of the
drum until the galvanometer returns to zero, and take the
reading on the drum.
This series of operations .takes less than half a minute.
The cells containing the liquids are mounted side by side, and
are easily inter:changed by pushing the slide along an inch or so
until it clicks into the correct position.
If we assume that the light intensity remains constant
throughout the series of operations. then th!,! current give-n
by the indicating cell at the end of operation 4 is the same as fl.t
the end of operation :2 (since in each case it balances the output
of the compensating cell). The difference in the illumination
condition in the two cases is that in the second case the
intensity-reduction produced by closing down the aperture is
substituted for the reduction produced by the absorption of
the specimen. The ratio of the area of the partly' closed
aperture to that of the aperture when fully open is thus a
measure of the absorption of the liquid. Since both photo-
cells are affected alike by changes in the intensity of the
lamp, the reading is unaff~ted by changes occurring during
the series of operations.
The sensitivity of the instrument in detecting small differ-
ences between the absorption of two liquids is greatly in-
creased by the use of a suitable light filter, the filter being
such that the range of wave .1engths over which it transmits
includes only the range of wave lengths over which the
solution absorbs. The use of a filter als~. usually, simplifies
the relation between the readings on the instrument and the
concentra~ion of the sol,ution. A set of 5 pairs of filterS" is
supplied with the instrument. The all-glass cells are made of
Pyrex ,gla-ss and are either I cm. or 4 cm. in length; the latter
are employed for feebly coloured solutions.
It should be noted that the photo-electric cells under
prolonged illumination with red light tend to behav.e irregu-
larly. For this reason it is pest, as a general rule, only to
switch on the instrument immediately before the start of a
602 Quantitative I norgamc A n{(l)'s~s
series of readings and to switch it off immediately when
finished. This precautien is really only necessary when using
red or yellow filters.
For the routine use of the absorptiometer in colorimetric
determinations, it is necessary to prepare a calibration curve
by taking readings. with a number 'of coloured .solutions of
known' concentration covering the required range. This
calibration curve remains valid so long as appreciable changes
do not take place in the spectral sensitivity of the photp-
electric cells, or the colour of the filter. The changes in the
cell and in the glass of the filters are generally very gradual
and the calibration curve need be checked only at wide
intervals, say, every 6 months. ...
The advantages of the" Spekker" absorptiometer' are :
(a) It is run directly from the electric I,llains supply-no batteries
are required.
(b) The readings are independent of the fluctuations of the mains
supply.
(c) The photo-cells are of the" rectifier" (or .. Sperrschicht ")
type, and are extremely robust and durable.
(d) The scale of the instrument is approximately linear with the
concentration of the solution.
(e) The instrument readings are not affected by variations in the
sensitivity of the cell or of the galvanometer, since a null methQd is
employed.
(f) The galvanometer which indicates the photo-electric current
is a robust pointer instrument and is used as a null indicator.
(g) Readings can be taken with as little as 7 m!. of liquid with a I
cm. cell; cells of -4 cm. and 20 em. l~gth are also available for use
with pale-coloured solutions.
V,5. Turbidimetry and nephelometry.~A coloured material
in true solution may be determined by comparison with the
colour of a standard-this is the true colorimetric method.
The transmission of light is measured and its 'reflection is
zero. With a colloidal dispersion of a coloured precipitate,
however, transmission and also some reflection of coloured
liglit occurs. If it can be assumed that the reflection is
negligible, then yolorimetric methods 'of determination may be
applied; here we have the phenomenon of turbidimetry, and
the instrument is used as a turbidimeter. If the reflection is
very great, colorimetric methods are subject to serious errors:
in such :t case the reflection of the light is measured. The
process of measurement is then c~lled nephelometry, and the
instrument is termed a nephelometer. With white precipitates,
Colorimetric Analysi: 693
colorimetric procedures cannot be applied and it is necessary •
to resort to thQse dependent upon. the determination of the
reflected light.
Tt is outside the scope of this volume 'to give a detailed
account of nephelometric apparatus and methods; for these
the reader is referred to special treatises on the subject (see
Appendix, Section A. 3). It may, however, be mentioned that a
Dubos<;q colorimeter may be readily modified for use as a
nephelometer; a special nephelometer attachment is also
available. The usual cups or tubes are replaced with clear
glass tubes with opaque bottoms; a powerful artificial light
source is employed and the light shining through ~he side of
the standard and unknown sample tubes is Ifeflected upward
through the tube and eyepiece. The amount ot light is
controlled as in colorimetry by variation of the depth of liquid
,under the plunger. Beer's law is rarely applicable in nephelo-
metry.
V. 6. Determination of the pH of solutions by colorimetric
methods.-A. Buffer solution method. A series of buffer
solutions (Section I, 20) is ,selected, differing successIvely in
pH by about 0'2, covering the pH range of the solutions under
examination. An approximate determination of the pH of
the liquid is made by the use of a multiple range indicator,
e.g., by the B.D.H. universal indicator, or by the systematic
use of a number· of indicators. This will give the PH of the
solution within 1 to 2 units, and will indicate the range of pH
of the buffer solutions required for comparison. Equal
volumes, say 10 mI., of the buffer solutions diht:ring succes-
sively in PH by about 0·2 are placed in a series of test tubes of
colourless glass, and having approximately the same dimen-
sions, and a small equal quantity of a suitable indicator for the
particular pH "range is added to each tube. A series of
differen! colours corresponding to the different pH values is
thus obtained. An equal volume (say 10 ml.) of the test
solution is treated with an equal volume of indicator to that
used for the buffer solutions, and the resultant colour is
compared with that of the coloured standard buffer solutions.
When a complete match is found, the test solution and the.
corresponding buffer solution have the same pH. So~etimes
a complete match is not obtained, but the colour of the test
solution falls between those of two successive standards, 'then
it is known that the pH value lies between those of the two
standards. Further buffer soh.ltiQns may then be prepared
694 Quantitative lttorganic Analysis
differing by 0·1 pH, if desired, and the pH value redeter-
mined. As a general rule, colorimetric methods cannot be
relied upon to give values of pH more accurate than to within
0·2 pH unit. For matching the colours, the buffer solutions
may be arranged in the holes of a test tube stand in order of
pH: the test solution is then moved from hole to hole until
the best colour match is obtained. Special stands and
standards for making the comparison are available commer-
cially. The commercial standards, prepared from buffer
solutions, are not permanent and must be checked every six
months.
For turbid or slightly coloured solutions, the direct com-
parison method given above can no longer be applied. The
,
, ,
• I
, interference due to the coloured
_.. 1_~ ._ .: _ substance can be eliminated in a
I I : simple way by a device due to
, : H. Walpole (1910). In Fig. 5-13,
A : C : A, B, C and D are glass cylinders
with plane bottoms standing in a
box which is painted a dull black
I !
on the inside. A contains the
I
, coloured solution to be tested
B: 0: (here the test solution + indica-
tor), B contains an equal volume
of water, C contains a solution of
,..-
I
~ ·II I known strength for comparison
• I I
(here the standard buffer solution
'C L!OH1--'}' + indicator), whilst D contains
FIG. 5-13. the same volume of the solution
to be tested as was originally added to A. The colour of the
unknown solution is thus compensated for.
For details concerning the preparation of buffer solutions,
see Appendix, Section A, 10.
B. Comparator (or permanent colour standard) method. In
this method comparison is made with a 'series of permanent
glass colour standards. Nine glass colour standards are fitted
into a disc, and the latter fits into a comparator, which is
furnished with four compartments to receive small test tubes
or rectangular cells, and is also provided with an opal glass
screen. The disc can revolve in the comparator, and each
colour standard passes in turn in front of an aperture through
which the solution in the cell (or cells) can be observed. As
the disc revolves, the value of the colour standard visible in
the aperture appears in a special indicator recess. The
Lovibond all-purpose comparator* is shown in Fig. 5-14:. A
FIG. 5-14.
(b)
FIG. 5-15.
similar apparatus is manufactured by F. Hellige and Co.t
(Fig. 5-15 (a) ), but this contains a prism attachment by
which the two colours being compared are brought side by
side in the eyepiece (Fig. 5-15 (b). The range of discs avail-
able with the Lovibond all purpose comparator include the
following:
.. Reproo,?ced by courtesy of The Tintometer Ltd .• MiHard, Salisbury,
England . .
t Supplied by A .. Gallenkarnp and Co. Ltd. , Sun Street. London. E.C.2
England, and by HeUige, Inc., Long Island City, N.Y .• U.S.A.
Qllantitative 1norgall~'c Analysis
, , --
I R,mge
pH
pH
.0·2--1-8 - - - - _ - --- - - - -
~ Range
- '
Cresol red (acid range) Bromo-thymol hlue 6-0- 7-6
»I-Cresol purple (acid range) 1·0---2·6 Phenol red . • 6·8- 8-4
Thymol blue (acid range) 1·2-2-8 Cresol r~d 7-2- 8·8
Bromophenol blue 2'8-4'4- Thymol blue 8·0-- 9·6
Bromo-cresol green 3·6--5-2 B.D.H. !JOIl indicator !Hl-U'O
RD.H. 4460 indi~tor 4·4--6·0 2 : 4-Dmitrophenol (a) 2·8-- 4-4
Methyl red 4·4--6·0 2 : 5-Dinitrophcnol (I') 4·0- 516
Chiaro-phenol red 4'8--6'4 12 ;-6-Dillitrophcnol ~(3) 2·g_,. 4·4
Bromo-cresol purple 5·2-6·8
It must be pointed out that the Lovi~ond comparator

utilises specially prepared B.D.H. indicators, whilst the
Hellige comparator must be employed with Merck's indicators.
Neither the discs nor the indicaiors are interchangeable in
the two instruments. AJl-approximate determination ofllthe
pH of the solution is first made with the B.D.H. universal
indicator (which covers the range 4·0 to n·O) and tpen a
suitable disc is selected_ Ten m!. of the unknown solution is
placed in the glass test tube or" cell, then the appropriate
quantity of indicator (usually 0·5 mi.) is added, :J.nd the colour
produced is matched against the glass disc. Provision is
made for the application of the Walpole technique-by the
insertion of a "blank" containing the coloured ~olution.
Results accurate to 0·2 pH can be obtained_
Mention should be made of the fact that a comparator lamp
is available for the Hellige comparator, thus permitting. deter-
minatipns in artificial light. A daylight lamp may also -be
employed with the Lovibond all-purpose tomparator.
Both compara:Jors may also be employed for the colori-
metric determination 'of small amounts of certain elements
and radicals, and special colour'discs are supplied for this
Eye Eye purpose (compare \:he B.D.H. Lovibond
nesslerimeter in Section V, 4A; this
instrument is more suitable for very
dilute solutions of the constituents under
D
determination)_ .
C. With the hydrogen ion colorimeter.
This is a modified form of the Duboscq
colorimeter.which utilises a new principle,
due toL. J.Gillespie (1921). It is illustrated
diagrammatically in Fig. 5-16. The vessels
fIG. 5-](;. A, B, C, D and E aI;e of colourless glass
with optically plane and paralld bottoms. A and C are fixed,
whilst B may he moved up and down. The position of B
is indicated on a scale, the zero graduation of wpich eorres-
ponds to the- position of B when Band C are in contact, and
the 100 mark of which corresponds to the position of B when B
. is in contact with.A. If a dilute solution of the acid form of an
indicator is pla~ed in Band t solution of the same concen-
tration of indicator but trans{ormed completely into the
alkaline form be placed in C, then it is clear that the position of
B will determine the ratio of the two forms of the indicator
which will be in the field of view. The unknown solution js
placed in E together with the same concentration of indicator
as in Band C. The auxiliary cups A and lJ'can be used for
compensation by the Walpole technique. B is then moved
until the two fields are matched; in this movement the com-
bined depth of the acid and alkaline colours remains the same
but their ratio is changed. The PH is then.calcul;:tted with
the aid of a simple formula based u'pon equation (xiii) in
Section 1,27 or can be read from tables. With a good instru-
ment, such as the Bausch and Lomb hydrogen ion colorimeter,
which embodies the Gillespie principle, but carried to a point
of refinement, readings may be duplicated to within 0·03 pH
units.*
Reference must also be made to the determination of PH,
to within O·()l to 0·04 pH units, with the Zeiss Pulfrich photo-
meter and with the Hilger" Spekker .. photo-electric absorp-
ti~meter. Further details will be found in the pamphlet-s
issued by the manufacturers ot these instruments.
V, 7. Some general remarks upon colorimetric determina-
tions.-.-As a general rule, the maximum error of a visual
colorimetric determination is not less than ± 2 per cent and
may be as high as ± 8 per cent or more, depending upon the
colour response of the observer's eye, the sensitivity of the
colour reaction and the method employed in colour matching.
With photo-electric equipment and also with the Zeiss Pul-
frich photometer it is quite possible to reduce the maximum
error of a colorimetric determination to less than ± 0·5 per
cent. 'When determining a minute quantity of a substance,
e.g., 0·01 mg., the error of the visual method of colour measure-
ment, even though it may be as high as ± 10 per cent, would
be of little practical account and hence photo-electric methods
• For further details, see the pamphlet issued by the Bausch and Lomb
Optical co.; also Snell and Snell, Colorimetric Methods of Analysis, 1936, p.
103 (Chapman amI EaU).
would possess no real advantage exc~pt speed of determina-
tion. When, however, a colorimetric method is employed for
determining a major constituent, such as copper in brass, the
increased accuracy obtainable by the use of a photo-electric
comparator (or Zeiss Pulfrich photometer) may give the
colorimetric determination a status equal to that of a volu-
metric or gravimetric method of analysis.
Reactions, suitable for direct colorimetric determination,
which do not suffer interference from other substances-at
least to a slight extent-are relatively few. Interference may
be a voided by :
(a) Suppression of ·the action of the interfering substance
by the formation of complex ions.'
(b) Removal 'of the interfering substance by extraction
with an organic solvent, e.g., iron may be removed from lead
by conversion into ferric chloride and extraction with ether,
or by conversion into "ferric thiocyanate" and extraction
with amyl alcohol and ether. _
(c) Isolation of .the substance to be determined by the
formation of an organic complex. which is then removed by
extraction with an organic solvent. This method may be
combined with (a) in which interfering metals are prevented
from forming soluble organic complexes by formation of a
complex ion and are thus left behind in the aqueous layer.
, (d) Separation by volatilisation. This is a good method in
the few c.ases where it is applicable, e.g., the distillation of
arsenic as the trichloride to seRarate it from other elements
prior to .its determination -b)1 the Gutzeit method.
(e) If the above methods cannot be used, it may be neces-
sary to isolate the substance by the ordinary methods of
inorganic analysis; double precipitation is frequently' neces-
sary to avoid errors due to occlusion and coprecipitation.
A general survey of the various methods suitable for colori-
metric determinations has been given in Sections V, 3 to V, 5.
In ·the following Sections reference will only be made to the
methods requiring the simplest of apparatus (colorimetric
titration, standard series, etc.), since such simple apparatu.s
will be available in all laboratories. It should be an easy:
matte); for the student to adapt the selected procedures to the
use of other-and more accurate-apparatus. Details for
the determination ~f a number of typical elements and radicals
are collected in the following pages; the list is by no means
exhaustive, but will serve to indicate the modern trend of
colorimetric methods.
V, 8. Determination of ammonia.-Discussi01£. J. Nessler in
1856 first proposed an alkaline solution of mercuric iodide in
potassium iodide as a reagent for the colorimetric determina-
tion of ammonia. Various modifications of the reagent have
since been made. When Nessler's reagent is added to a dilute
~ammonium salt solution, the liberated ammonia reacts with
the reagent fairly rapidly but not instantaneously to ·form an
orange-brown product, which remains in colloidal solution, but
flocculates on long standing. The colorimetric comparison
must be made before flocculation occurs. The product has
the formula NH2Hg2Ia (M. L. Nichols. and C. O. Willets,. 1934).
The reagent is employed for the determination of ammonia in
very dilute ammonia solutions and in water. In the presence of
interfering substances, it is best to separate the ammonia first by
distillation under suitClble conditions (compare Section Ill, 2O).
The method is also applicable to the determination of nitrates and
nitrites: these are reduced in alkaline solution by Devarda's alloy
to ammonia, which is removed by distillation. The procedure is
applicable to concentrations of. ammonia as low as 0·1 mg. per litre.
Nessler's reagent is prepared as follows. Dissolve 50 g. of
potassium iodide in about 35 ml. of cold ammonia-free water.
Add a saturated solution of mercuric chloride until a slight
precipitate appears. Add 400. m!. of clear 9N sodium or
potassium hydroxide~ Dilute to 1 litre, allow to stand, and
decant from the sediment. Keep stoppered and in a da~
place.
Ammonia-free water may be prepared in a conductivity
water stilI (compare Section n, 10C) ar as follows. Redistil
.500 mL of distilled water in a Pyrex apparatus from a solution
containing 1 g. of potassium permanganate and 1 g. of anhy-
drous sodium carbonate; reject the first 100 m!. portion of
the distillate and then collect about 300 m!.
»Procedure. For practice in this determination, the student
may employ either a very dilute ammonium chloride solution
or ordinary distilled water which usually contains sufficient
ammonia for the exercise.
Prepare a standard ammonium chloride solution as follows.
Dissolve 3·141 g. of ~.R. ammonium chloride, dried at 100°C,
in ammonia-free water and dilute to I litre with the same
water. This stock solution is ,too concentrated for most
purposes. A standard solution is made by diluting 10 ml. of
this solution to 1 litre with ammonia-free water: 1 m!.
contains 0·01 mg. of NH~.
700 Quantitative I norgq,_nic Analysis
If necessary, dilut~ the sample to give an ammonia concen-
tration of I mg. per litre (Hehner cylinders, Fig. 5~5, are useful
for this dilution), and fill a 50 ml. Nessler tube to the mark.
Prepare a ·series of Nessler tubes containing the following
volumes of standard ammonium chloride solution diluted to
50· ml.: 1'0, 2'0, 3'0, 4'0, 5·0 and 6·0 ml. The standards..
contain'O'OI mg. of NHa for each ml. of the standard solution.
Add 1 mI. of Nessler's reagent to each tube, allow to stand for
10 minutes, and'compare the unknown with t.he standards in a
Nessler stand (Fig. 5-2) or in a nesslerimeter (Fig. 5-3). This
will g~ve an approximate figure which will enable another
series of standards to be prepared and more accurate results
obtained (compare Section V, 4A).
V, 9. Determination of nitrites.-Discussion. When acetic
acid solutions of sulphanilic acid and a-naphthylamine are
acted upon by nitrous acid, a red colouration is produced
which may be use,d for the colorimetric..determination of small
amounts of nitrites. * The sulphanilic acid is converted into
the corresponding diazo compound which couples with the
a-naphthylamine to form a-naphthylamine-p-azobenzene-p-
sulphonic acid, a red azo dye. The full colour ~does not appear
for several hours, but a satis~actory colour comparison can be
made after 5 to 10 minutes, provided the sample and standard
are treated in the same way.
The r~ent is prepared as follows:
Solution 1. Add 1 g. of A.R. sulphanilic acid to 14·7 g.
of glacial acetic acid. and 15 mI. of water. Warm until
solution is complete, dilute with 300 ml. of water, and stir
during the additio~
Solution 2. Add 0·2 g. of A.R. a-naphthylamine to 14·7 g.
of glacial acetic.acid and 25 ml. of water. Warm until solu-
tior~ is co~plete, dilute with 300 mI. of water, and stir during
the dilution.
Mix equal volumes of the solution_s immediately before use.
Standard nitrite solution. Dissolve 4·f)3 g. of A.R sodium
nitrite in water and dilute to I litre. Dilute 10 ml. of this
stock solution to 1 litre: 1 ml. of tq,is sta,ndard solution
corresponds to 0·01 mg. of nitrite nitrogen.
* The colour produced by a-di~ethyl-amino-naphthalene is superior in
intensity. stability and brilliance to that obtained with a-naphthylamine;
it is similarly used to the latter, except that a solution is obtained by dissolu-
tion in a solution of 4N acetic acid in 95' per cent methyl alcohol. Giblin
(1936) has proposed dimethylanilinc.
Colon'metric Analysis 701
Procedure. For pra:ctice in this determination either a very
dilute solution of sodium nitrite or ordinary distilled water
may be used. If necessary, dilute the sample to give a nitrite
concentr.ation of 1 mg. per litre. Prepare a series of Nessler
tubes containing the following volumes of the standard nitrite
solution diluted to 50 m!. : 0·1, 0·2, 0·4, 0·6, 0·8 and 1·0 ml.
Place 50 ml. of the unknown solution in a Nessler tube. To
each tube add 2 mI. of the reagent, and stir with a special glass
rod (see Section V, 4B). After 10 minutes, compare the
'colours of the tubes. If a closer approximation is required,
prepare a fresh series of standards on either side of the first
value and repeat the comparison, starting all the tests at the
same time.
V, 10. Determi.J;lation of iron.-Discussion. Two methods
will be describ~d: the thiocyanate method and the thiogly-
collic acid method.
A. Thiocyanate method. Ferric iron reacts with thiocyanate
to give an intensely-red coloured' c;ompound which remains in
tn!e solution, whereas ferrous iron does not react. The
colour is usually assumed to be due to ferric thiocyanate:
Fe+++ + 3CNS- ~ Fe(CNS) 3,
but the recent work of H. I. Schlesinger and H. B. V. Valken-
burgh (1931) has shown that the red solution contains the
complex [Fe(CNS)sr - -: ion which is red.
Fe+++ + 6CNS- ~ [Fe(CNS).r --
A large excess of thiocyanate should be used since this in-
creases the colour intensity. Strong acids (hydrochloric or'
nitric acid--co1'l.centration 0·5 to O·5N) should be present to
suppres~ the hydrolysis:
Fe+++ + 3H 20 ~ Fe(OH)3 + 3H+.
Sulphuric acid is not recommended because sulphate ions have a
certain tendency to form complexes with ferric ions. Silver,
copper, nickel, cobalt, and large quantities of mercury (> I g. per
litre), zinc, cadtnj,um and bismuth (> 0-1 g. ~er litre) interfere.
Mercurous and stannous salts, if present, should be converted into
the mercuric and stannic salts, otherwise "the colour is destroyed .
.Phosphates, arsenates, fluorides, oxalates and tartrates interfere
sil}-ce they form fairly stable complexes with ferric ions; the in-
fluence of phosphates and arsenates is reduced by the presence of a
comparatively high concenl:ration of acid. _.
When large quantities of interfering substances are present, it is
usually best to proceed in either of the following ways: (i) remove
the'iron by precipitation with a slight excess of ammonia solution,
and dissolve the precipitate in dilute hydrochloric acid; (ii) extract
the" ferric thiocyanate" three times t:ither with pure ether or,
better, with a mixture of amyl alcohol and pure ether (5 : 2) and
employ the organic layer for the colour comparison. *
B. Thiogiycollic acid method. Thioglycollic ac{d when
added to a solution of a ferrous salt and the mixture rendered
alka~ine with ammonia solution, produces an intense red
colouration. Ferric salts in concentrations greater than 1 in
100,000 yield a transient blue colouration; this is imme-
diately reduced to a colourless ferrous thioglyeollate (I),.
which then yields the intensely coloured ferrous thioglycollate
ion (III) in amIl}oniacal solution (the thioglycollic acid is
oxidised to dithioglycollic acid (II), which gives no colour
with either ferrous or ferric iron) :
2Fe+++ + 2HS.CH 2 .COOH --!> 2Fe++ + (S.CH 2:COOH)2 (II)
+2H+.
Fe++ + 2HS.CH 2 .COOH ~ Fe(S.CH 2.COOH)2 (I) ~
Fe(S.CH2.COO)2 - - (III).
Strong alkalis slowly discharge the colour. A method is thus
available for determining both ferrous and ferric iron colori-
metricalJy. The ferric iron is determined-with thiocyanate;
the colour is stable in acid solution and is discharged by
alkalis. '];.he total iron is determined with thioglycollic acid,
- and the ferrous iron by difference.t
Fluorides, phosphates, arsenates, sulphates, etc., are without
effect. Chromium, nic~el, cobalt, manganese and uranium inter-
fere, whilst. other common metals if present in the order of the
. concentration of the iron have-no effect. Aluminium is rendered
inn.oc\louS by the addition o~ excess of citric aCidr.or tartaric acid.
Silicates (which have a bleaching effect on the colour) and oxidising
agents must be absent. .
The thioglycollic acid method is suitable up to a concen-
tration of 1 part in 5,000,000 and the thiocyanate method to
several parts per million.
Procedure (thiocyanate method). Prepare the following
solutions : ~
1. Standard solution' of ferric iron. (a) Dissolve 0·7022 g.

of A.R. ferrous ammonium sulphate in.lOO ml. of water, add
5 ml. of 1 .: 5 sulphuric acid, and run in cautiously a dilute
solution of potassiuVJ, permanganate (2 g. per litre) until a
• For experimental details of this process, see Lundell, Hofmann and Bright,
Chemical Analysis of I'Yon and Steel, 1931, p. 44; see also Section I, 7lB. .
t Ferrous iron may also be determined directly with dimethylglyoxime;
a red colour is produced.
Colorimetric Analysis 703'
slight pink colouration remains after stirring well. Dilute
to 1 litre and mix thoroughly. 1 Ml. === 0·1 mg. of Fe.
(b) Dissolve 0'864 g. of A.R. ferric ammonium sulphate in
water, add lQ-ml. of concen~rated hydrochloric acid and dilute
to llit~e. 1 Ml. == 0·1 mg. of Fe.
2. Potassium thiocyanate solution. Dissolve 40 g. of A.R.
potassium thiocyanate in 100 ml. of water; the solution is
ca.4N.
Dissolve a weighed portion of the substance in which the
amount of iron is to be determined in a suitable acid, and
evaporate nearly to dryness to expel excess of acid. Dilute·
slightly. with water, oxidise the iron to the ferric state with
dilute potassium permanganate solution or with a little
bromine water, and make up the liquid to 500 ml. or ot,her
suitable volume. Alternatively, use the iron solution pro-
v,ided by the teacher and oxidise with bromine water: Place
59 ml. of tbe solution in a Nessler cylinder, and add 5 ml. of
the thiocyanate solution and 2 to 4 ml. of 4N nitric or hydro-
chloric acid. Add the same amounts of the reagents to 50 m!.
of water contained in another similar Nessler tube, and run
in the standard iron solution from a burette (use a long glass
tube w~th bulbed or flattened end for mixing) until the colours
are m~tched; note the exact volume (x m!.) of the standard
iron solution added.' Repeat the determination using (50-x)
ml. of water. Comparison of the standard and unknown
should be made soon after preparation since the colour fades
on standing.
Procedure (thioglycollic method). The unknown solution 1s
first dilutM so as to have a concentration of iron of about 1
part in 500,000. This is be~t done by testing the.unknown
solution qualitatively: after a little experience, the intensity
of the colour produced will indicate the proportion for d~lution.
Place 5 ml. of the solution in a test tube graduated at 5 m!.
intervals, add l"Or 2 drops of thioglycoHi.c acid and 0·5 m!.
of concentrated ammonia solution. Compare the colour pro-
duced with the colours of a series of standards of known
concentr:ation of iron which have been similarly prepared.
The Lovibond all-purpose comparator is very suitable for this
determination (compare Section V, 5B).•
V, 11. Manganese.-Discussion. Small quantities of man-
ganese are invariably determined colorimetrically by oxida-
tion to permanganic acid. :rhe oxidising agents which have
been employed include lead dioxide a.nd strong nitric acid,
704 Quantt'tative Inorganic Analysis
sodium bismuthate, potassium or ammonium persulphate in
the presence of a little silver nitrate as catalyst, and potassium
periodate. The first two are not very suitabl~ since the
excess of the reagent must be removed by fi.ltration. The
persulphate metbod, although widely used, is generally un-
satisfactory for, in a number of cases, the reaction proceeds
incompletely to the permanganate stage. The period ate
method is free from these disadvantages. In hot acid solution,
periodate oxidises manganese quantitatively to permanganic
acid:
2Mn++ + 510 4 - + 3H 20'
•
.
= 2Mn0 4 - + 5103 - + 6H+ .
When ready for test the solution should not contain more • than 2
mg. of manganese per !OO mI., otherwise the colour will be too dark
and colour matching will be difficult. Frequently ferric iron is
added to the standard in amount equal to that found independently
to be present in the sample. Phosphoric acid must be present to
prevent the precipitation of ferric periodate and iodate and also~to
decolourise the ferric iron (by complex formation). If chlorides
are present, it is necessary to evaporate with a mixture of nitric
and sulphuric acids until fume:_; of the latter appear. Chlorides
react with the periodate.
Procedure. For practice in this determination, the man-
ganese content of a suitable steel (1, may be evaluated.
Weigh out accurately.a suitable quantity of the steel' (0'1 to
0·2 g. for steels containing up to 1 per cent of Mn) into a
conical flask, dissolve it in 20 to 50 ml. of 1 : 3 nitric acid, and
boil for 1 or 2 minutes to expel oxides of nitrogen. ' RemQve
from the burner, and add 0·5 to 1·0 g. of A.R. a.mmonilim
persulphate; boil for 10 to It> minutes to oxidise .carbon
compounds and to destroy the excess of persulphate. If any
permanganate colour develops or oxide~ of manganese separ-
ate, add a few drops of sulphurous acid or sodium sulphite
solution to reduce the manganese and render the solution clear,
and boil for a few minutes to expel the excess of sulphur
dioxide. Dilute the solution to ca. 100 ml., add 5 to 10 rill. of
A.R. syrupy phosphoric acid and 0·5 g. of potassium perio-
date (2): boil for 1 minute' and keep hot for 5 to 10·minutes.
Cool the solution and make up to 250 ml. in a volumetric
flask and mix thoroughly. For a comparison solution, use
either a steel of known manganese content which has been
similarly treated (3) or a standard solution containing 4·55 ml.
of O'IN potassium permanganate diluted to 250 ml. (1 m!. ==
0·02 mg. Mn). Any of the methods described in Section V, 4
may be used for the colorimetric determination; that utilising
a Duboscq colorimeter is particularly suitable.
Notes. 1 Ridsdale's" Mild Steel, No.1" (one of the Analysed Samples for
Students) or the Bureau of Analysed Samples I ' Carbon Steel, ' E ' or J R ' "
or British Standardised Steel Sample, No. 11 (supplied by the N.P.L.) may be
employed. Suitable steels are also available from the U.S. Bureau of
Standard; see Appendix. Section A. 8.
2. Each 0·1 g. of Mn requires 1 g. of potassium periodate.
3. Such standards. containing excess of period ate. are stable for 2 to 3
months.
V, 12. Titanium.-Discussion. Hydrogen peroxide pro-
duces with an acid solution of titanjum sulphate a yellow
colour, which js said to be due to pertitanic acid. With
small amounts of titanium the intensity of the colour is pro-
portional to the amount of the element present, and therefore
forms the basis of a colorimetric' method for the determination
of titanium. The method is applicable to solutions contain-
ing up to 0·1 mg. of Ti0 2 per ml. Comparison is usually made
with standard titanic sulphate solutions; two methods for
their preparation (from potassium titanium oxalate and
potassium fluotitanate) are described below. The hydrogen
peroxide solution should be of about 3 per cent strength
(1 O-volume) , and the solution should contain at least 5 per
cent by volume of sulphuric acid in order to prevent hydrolysis
to a basic sulphate and to prevent condensation to meta-
titanic acid. The colour intensity increases with rise of
temperature, hence the solutions to be compared should have
the same temperature, preferably 20° to 25°C.
Elements which interfere are: (a) iron, nickel, chromium, etc.,
because of the colour of their solutions; (b) N'anadium, molybdenum,
cerium, and, under some conditions, chromium, because they form
coloured compounds with hydrogen peroxide; (c) fluorine (even in
minute amount) and large quantities of phosphates, sulphates and
alkali salts (the influence of the last two is largely reduced the greater
the concentration of sulphuric add present-up to 1'0 per ceni).
The influence of elements of class (a) is overcome, if present in small
amount, by matching the colour by the addition of like quantities
of the coloured elements to the standard befoH! hydrogen peroxide
is added. Whe)) large amounts of iron are present, as in the analysis
of cast irons and steels, two methods may be adopted: (i) phos-
phoric add can be added in like amount to both unknown and
standard, after the addition of hydrogen peroxide; (ii) the iron
content of the unknown solution is determined, and a quantity of
standard ferric alum solution, containing the same amount of iron,
is added to the standard solution. Large quantities of nJckel,
chromium, etc., must be removed. Elements of class (b) must als,o
AA
.
be removed; vanadium and molybdenum are most easily separated
by precipitation of the titanium with sodium hydroxide solution in
the pr~sence of a little iron; cerium may be elimipated 'as oxalate.
Fluorine h<;!S the most powerful effect in bleaching the colour: it
must be removed by repeated evaporation with concentrated sul-
phuric acid. The bleaching effect of phosphoric acid is overcome
by adding a like amount to the standard, or by adding 1 mI. of 0·1
per cent uranyl acetate solution for each 0·1 mg. of Ti present.
Procedure. The experimental details. for the determination
of titanium in a silicate rock (e.g., felspar) and in a cast iron
or steel will be given.
Preparation of standard titanium solution. From potassium
titanittm oxalate. Recrystallise the commercial product onc.e
from water, if necessary, and dry at room temper~ture for
several days. Transfer 4·5 g. of pure KjlTiO(C z0 4)2,2H 2 0 to
a Kjeldahl flask (Section m, 20), add 8 g. of ammonium sul-
phate and 50 ml. of concentrated sulphuric acid, and heat the
mixture gradually. Finally boil for 10 minutes to destroy
the oxalic acid. Cool, pour the solution into 750 m!. of water,
and dilute to 1 litre. Determine the exact titanium content
by witlidrawing 50 ml. of the solution, then. dilute to 200 mI.,
heat to boiling, and render faintly ammoniacal. Contin\\e
the boiling for 2 to 3 minutes. Filter off· the ;precipitated
Ti0 2 ,xH zO on a filter paper, wash with hot water until free
from alkali salts, ignite moist in the filter, and weigh as Ti0 2 •
Carry out the determination in duplicate. Calculate the TiO z
content per m!. of solution.
From potassium titanium fluoride. If necessary, recrystal-
lise the potassium fiuo-titanate fron! water, preferably in a
platinum dish, and dry at :L05°C; the salt then has the for-
mula K2TiF6,H20. Weigh oilt 1·62 g. of the pure mono-
hy<;lrate into a platinum dish, add 50 mI. of 1 : 1 sulphuric
jil.cid, and evaporate nearly to dryness. Cool, wash down the
sides of the dish, add more sulphuric acid and again evaporate
until most of the sulphuric acid has been removed. Repeat
the operation again in order to insure that all hydrogen
fluoride is expelled. Cool, transfer quantitatively to a 500
m!. volumetric fiask, add sufficient sulphuric acid so that its
final concentration in the solution is at least 5 per cent by.
volume, make up to the mark and mix well. If the initial
salt were perfectly pure, thtr solution would contain 1·00 mg.
of TiO z per ml. It is best, however, to determine the exact
t~tanium content as above.
Colorimetric Analysis 707 -
A. Titanium in felspar. Dissolve a known weight of the
" mixed oxides" (Section IV, 95, 03-analysis of a felspar)
in sufphuric acid, transfer to a 100 mL flask, add sufficient
sulphuric acid so that the total sulphuric acid content is
ultimately 10per cent, and dilute to within 10 mL of the mark.
If the solution is coloured because of the presence of iron, add
suflicient phosphoric acid to decolourise it, and make up to
the mark. Prepare a standard solution by placing 80 ml. of
10 per cent sulph"Qric acid in a 100 ml. standard flask, add the
same amount of iron (as ferric alum solution) as is present in
the un~nown, decolourise by the addition of phosphoric acid
dropwise, at}d make up to the mark. Transfer 50 mL of
each solution to Nessler cylinders~ add 3 ml. of 10-volume (3
per cent) hydrogen peroxide to each and mix by means of a
long glass tube with a bulb or flattened end. Now add the
standard titanium solution from a burette to the blank until
after mixing the colours match; note the volume of standard.
titanium solution employed and a,dd this volume of 10 per
cent sulphuric acid to the unknown solution. This will give
an approximate va,lue of the titanium content of the solu-
tion. More exact values can then be obtained by comparison
in a Duboscq or other colorimeter.
B. Titanium in cast iron. Weigh out accurately about 0·5

g. of the cast iron into a 100 ml. . conical flask, add 6 ml. of
1 : 3 sulphuric acid, and heat gently until solution is complete;
if a cake of solid separates, dissolve it in a few ml. of water.
When solution is complete,:" cool somewhat, and add 10 ml.
of 1 : 1 nitric acid. Heat until fumes of sulphur trioxide are
evolved, and cool to room temperature. Transfer the solu-
tion to an accurately graduated 25 ml. test tube or flask, rinse
the flask 2 or 3 times with a little dilute sulphuric acid
(1 : 20), and add water to the 25 ml. mark. Filter through a
dry filter paper into a 30 ml. test tube: do not wash the tube
or the paper.
Take an aliquot portion (10, 15 or 20 ml.) of the clear
filtrate, and add syrupy phosphoric acid dropwise until the
solution is decolourised. Add 3 ml. of 10-volume (3 per cent)
peroxide, mix, and compare in a colorimeter against a stan-
.. An insoluble residut:' of 5Hica can be " fumed" off with hydrofluoric acid,
followed by treatment with concentrated sulphuric acid to remove excess of
the former acid. Large quantities of iron (e.g., in fer.ro-silicon, etc.) are best
handled as ferric chloride by extraction with an organic solvent prior to the
titaniulll determination.
dard sample of iron* which'has been treated exactly as the
unknown sample. The simplest procedure is to add to the
tube of lighter colour a standard solution of titanium -from a
micro-bu·rette and to the other add similar amounts of 1 : 20
sulphuric acid until the colours match when the solutions are
\VeIl mixed. The percentage of titanium may then-be calcu-
lated from the titanium content of the iron and the titre of
the standard titanium solution .
.V,13. Vanadium.-Discussion. When hydrogen peroxide
is added to a solution containing small quantities of
vanadium (up to 0·1 mg. of V per ml.) and 15 to 20 per cent
of its volume of sulphuric acid, a reddish-brown colouration
is produced; this is usually attributed to the formation of
" pervanadic acid." This forms the basis of the colorimetric
procedure for the determination of vanadium. The method
is not so. satisfactory as for titanium. The colouration is
intensified by strong acids and a large excess of hydrogen
perpxide tends to bleach the colour, hence the experimental
conditions must be rigorously controlled. The vanadium
must be in the quinquevalent state.
The colour is unaffected by moderate amou~ts of hydrofluoric
acid; this acid is therefore sometimes employed to prev«nt inter-
ference by small quan. ·ties of titanium. The method is not appl!c-
able, in the presence of elements which give colourations with
hydrogen peroxide, such as titanium, molybdenum and cerium;
tungsten also interferes and must be removed. Jron can be de-
colourised by the addition of either phosphoric or hydrofluoric
acids. In the presence of elements· which yield coloured solutions,
such as chromium or nickel, it is best to add equal amounts of these
elements to the standard solution. If steel is being anplysed, the
most convenient procedure is to use a like steel as standard.
Procedure (vanadium in steel). Prepare a standard solution
oj vanadium by dissolving 1·146 g. of A.R. ammonium vana-
date NH,VO a in water and making up to 1 litre; 1 m!.
contains 0·5 mg. of V.
Dissolve 1 g., accurately weighed, of the steel in 50 m!. of
1 : 4 sulphuric acid. When solution is complete, introduce
10 ml. of concentrated nitric acid, and boil until nitrous fumes
are no longer evolved. Dilute the solution to 100 m!. with
hot water, heat to boiling and add saturated potassium per-
manganate solution until a pink coJour persists or a precipi-
• Such standards may be obtained from the Bureau of Analysed Samples.
Middlesbrough. England. or from the U.S. Bureau of Standards, Washington:
tate is formed.'" Boil for 5 minutes. Filter off on asbestos
any tungstic oxide or manganese oxide which may be precipi-
tated. Add a slight excess of freshly prepared sulphurous
acid, and boil off the excess. Cool, add 5 ml. of syrupy
phosphoric acid and 5 m!. of 10-volume hydrogen peroxide .•
Simultaneously with the first decomposition prepare in an
analogous manner a comparison solution from a standard steel
which contains no vanadium, but is otlierwise similar. Alter-
natively, prepare a comparison solution from a plain carbon
steel and add appropriate amounts of solutions M chromium,
nickel, etc., if the unknown contains these elements.
:I'ransfer the two solutions to 100 ml. Nes~ler cylinders,
and treat the 'blank with standard ammonium vanadate
solution, added from a burette, until the colours match.
Allow 2 minutes for the development of the colour. A more
precise comparison may be made with a Duboscq colorimej;er
or with the other apparatus of Section V, 4.
V, 14. Chromium.-Discussion. Small amounts of chro-
mium (up to 0·05 per cent) may be determined colorimetric-
ally by comparing the colour of chromates in alkalin~ solution;
uranium interferes. A standard solution of chromium is
prepared by dissolving 0·2555 g. of A.R. potassium chromate
and 0·1 g. Of J\.R. sodium carbonate in water and diluting to 1
litre; 1 m!. corresponds to 0·1 mg. Cr2 0 3 •
A better method is' to employ diphenylcarbamde
CO(NH.NH.CeH o)2 (Cazeneuve's reagent, 1900). In acid
solution chromates form a soluble violet compound with this
reagent; the test is sensitive to 1 part in a million or better.
Cadmium, magnesium, mercury, molybdenum, copper, lead,
silver, gold, zinc, nickel and cobalt interfere; permanganates,
if present, may be removed by boiling with a little alcohol.
Procedure. Prepare the diphenylcarbazide reagent by dis-
solving 1 g. of diphenylcarbazide in 100 ml. of alcohol con-
taining a trace of sulphuric acid. The standard chromium
solution is made as detailed in the Discussion.
Faintly acidify a definite volume (x ml.) of tpe liquid con-
taining the chromate (0'02 to 0·5 mg. of Cr) with sulphuric
acid, add 2 ml. of the reagent and 20 ml. of 1 : 1 sulphuric
acid. Match the colour against that obtained by adding the
• Alternatively, add 20 ml. of 1 per cent silver nitrate solution, heat to
boiling, and add slowly, while stirring, a concentrated solution of ammonium
persulphate until a permanent pink colour develops. If no pink colour is
obtained, add a few drops of 2 per cent potassium permanganate solution.
Boil for 5 minutes.
710 Quantitative Inorganic A natysis
standard chromate solution from a micro-burette to a mixture
of x m!. of water, 2 m!. of the reagent, and 20 mI. of 1 : 1
sulphuric acid (for. further details, see Section V. 4B).
, V, 15. Aluminium.-Discussion. Aluminium salts give a
'bright red lake when treated with the ammonium salt of aurin
tricarboxylic acid (" aluminon") in the presence of an acetic
acid-acetate buffer. ~f the concentration of the aluminium
is small (up to 0·05 mg. in 50 mI.), a coloured liquid (faint pink
. to deep red.accOrding to the concentration of the aluminium)
is obtained, which is suitable for the colorimetric determination
of aluminium.
The chief interfering element is iron-even minute amounts give a
purplish-red tint-and must be entirely removed:' this is most
simply effected by extraction as thiocyanate with ether-amyl
alcohol or with amyl alcohol alone (compare Section V. 9A). Bis-
muth, lead, antitnony, stannic tin, mercuric mercury, titanium and
siliCic acid give white precipitates. Large amounts of alkaline or
rare earths, of basic acetates or hydroxides of chromium, ce,rium,
thorium and zirconium give red lakes, and phospliates also interfere;
these are either dissolved or decolourised by the addition of ammo-
nium carbonate solution. Reducing agents must be absent.
Procedure. Prepare the aluminon reagent by dissolving
0·2 g. of ammonium aurin tricarboxylate in 100 ml. of water.
Prepare also the standard aluminium. solution hy dissolv-
jng 1·759 g. of A.R. potassium aluminium sulphate
K 2SO"A1 2 (SO,) 3,24H 2 0 in water and diluting to 1 litre ; 1 ml.
contains 0·1 mg. AI.
Prepare 20 to 50 ml. of a clear, dilute hydrochloric acid
(2 parts of concentrated acid: 98 parts of water) solution that
is free from iron, from appreciable concentrations of salts that
interfere because of their colour, and, preferably, from sub-
stances, other than aluminium, that are precipitated by
ammonium carbonate solution. The solution should contain
not more than 0·05 to 0·1 mg. of AI. Add 10 ml. of 1 : 3
acetic acid, 5 ml. of the reagent, and finally a 10 per cent
solution of ammonium carbonate in 1 : 2 ammonia solution
slowly and :with constant stirring until an excess of 5 to 10
ml: is present. Compare the colour produced with standards
containing known amounts of aluminium until a match is
obtained.
The following procedure, due.. to Scherrer and Mogerman
(1938), is superior to the above and is particularly well
adapted to the determination of small quantities' of aluminium
in non-ferrous alloys, such as solders, bearing metals, etc.
Colorimetric Attalysis 711
Interfering elements are completely removed by the use of
the mercury cathode cell shown in Fig. 4-] 3 (c); here the anode
consists of.a platinum wire, I mm. in diameter, which projects
about 12 cm. into the solution, and the anolyte and mercury
are simultaneously vigorously stirred by the use of a stirrer
carrying two blades so situated that one is immersed in the
anolyte and the other in the mercury. With about 50 ml. of
mercury and a current of 3 amps" '1 g. of copper or iron can be
removed In about 45 minutes and 1 g. of tin, antimony, lead,
zinc, etc. in 2 to 3 hours.
Dissolve a convenient weight of the alloy in an appropriate
acid. Sulphuric acid is the best acid to employ, if P9ssible ;
if solution is effected with nitric acid, the latter must be
removed by gentle fuming with sulphuric acid: the final
content of sulphuric acid should not exceed 20 per cent by
volume. Dilute the solution to 25 mI., filter off any lead
sulphate which has separated, and wash the filtrate into a
mercury cathode cell and electrolyse. When electrolysis is
complete, separate the electrolyte from the mercury whilst
the current is kept on. The volume of the solution should
now be from 40 to 75 mI. Add 5 mI. of concentrated hydro-
chloric acid, 5 ml. of glacial acetic acid and 5 ml. of the alu:
minon reagent. Mix well whiJst cautiously adding concen-
trated ammonia solution until the solution becomes clear,
although still acid to litmus and still deeply coloured. Place
a pi~ce of litmus paper against the inner surface of the beaker,
and, while stirring constantly, add concentrated ammonia
solution at a rate of about 1 drop every 2 seconds until about
2 m!. has been added; then add 1 drop every 3 or 4 seconds
until the iitmus paper turns blue. Add 5 ml. of glacial
acetic acid, allow to stand for 5 minutes, and neutralise as just
described. Finally add 5 ml. of concentrated ammonia
solution. When the solution has cooled to room temperature,
compare the colour obtained with that of solutions of known
aluminium content which have been similarly prepared.
V, 16. Molybdenum.-Discussi'On. The method for the
colorimetric determination of up to 1 per cent of molyb-
denum, particularly in steels, is based upon the brown
colouration produced by the formation of molybdenum
thiocyanate in an acid medium. The brown colouration is
extracted by ether, butyl acetate or cyclohexanol-the last-
named is the best solvent-and the colour compared with
standards which have been similarly prepared.
Nickel, chromium, vanadium, silicon and copper do not interfere
in the percentages usually encountered in steel, although if the
copper content is greater than 0·13 per cent it is necessary to separate
the insoluble copper thiocyanate before estimating the molybdenum.
Procedure (molybdenum in steel). Prepare the following
reagents:
Sodium or potassium thiocyanate solution. Dissolve 10 g.
of A.R. sodium or potassium in 200 ml. of water. .
Stannous chloride solution. Heat 50 g. of stannous chloride
crystals with 40 ml. of concentrated hydrochloric acid until
dissolve¢l, and dilute with 240 ml. of water. Preserve over
metallic tin to prevent re-oxidation.
cycloHexanol. R~distil the purest commercial cyclohexanol,
and collect the fraction b.p. 158° to 160°C. Shake 50 ml. of
cyclohexanol with 5 ml. of the thiocyanate and 25 m!. of the
stannous chloride solutions iIi a separating funnel, and discard
the aqueous solution. This reagent must be prepared freshly
as required.
Standard molybdenum solution. Heat 0·71 g. of A.R.
molybdic acid (containing 85 per cent Mo0 3) with 100 ml. of
concentrated nitric acid and 30 m!. of concentrated hydro-
chloric acid until dissolved. Add 100 ml. of water, boil, cool,
and dilute to 2000 ml. The molybdenum content (about 0·2
mg. per ml.) may be computed from the weight of molybdic
acid employed, but is best determined in 50 ml. of solution as
lead molybdate (Section IV, 23A) •
. Dissolve 0·1 g. of the sample of steel (or iron) and also a
standard sample* of the same weight in 5 m!. of nitric acid,
sp. gr. 1'13, and 5 mt of 60 per cent perchloric acid (DANGER,
see Section IV, 70) in 150 ml. beakers provided with raised
clock glasses or with 'Fisher It'speedyvap'' beaker c,overs.t
Boil for 10 minutes until copious white fumes commence to be
evolved, cool, and dissolve in 15 ml. of water. Add 5 ml. of
the thiocyanate solution to ~ach beaker, and transfer to 100
m!. separating funnels, rinsing the beakers first with 10 ml. of
the stannous chloride solution and then with 10 or 15 ml. of the
cyclo hexanol. Shake the funnels for 1 plinute and reject the
• For example, either the U.S. Bureau of Standards Cr-Mo steel, No. 72a
or a similar molybdenum-free steel to which a known amount of the standard
molybdenum solution has been added.
t For stainless steels it may be necessary to use also 5 ml. of 1 : 1 hydro-
chlo.ic acid. .
;I.queous-solution. Add 5 ml. of the thiocyanate solution and
15 ml. of stannous chloride solution, and shake again for 1
minute. Draw off the wash solutions, and filter if an appre-
ciable quantity of insoluble cup~ous thiocyanate is present.
Transfer to similar colour comparison tubes, and add the
cyclohexanol saturated with thiocyanate and stannous chloride
solutions from a burette until the colours of the tubes are
.identical. The weight of molybdenum in the sample and
standard is then directly proportional to the vj)lumes (com-
pare Section V, 30).
V. 17,. Nickel.-Discussion. This element may be deter-

mined with dimethylglyoxime (compare Section IV, 12), but
since the nickel complex is sparingly soluble, only minute
amounts of nickel may be determined colorimetrically by this
process, for otherwise the complex precipitates. Ferrous
iron, which gives a red colour with the reagent, interferes and
must be removed; the interference of small amounts of
cobalt is prevented by the addition of excess of ammonia
solution. The sensitivity of the reaction is about 'I part in
500,000 parts of solution, but may be increased by the addition
of hypochlorite or hypobromite prior to_ the addition of the
reagent. The coloured complex formed after oxidation is,
however, quite soluble in water. Interference from 'small
amounts of iron and aluminium may be prevented by the
introduction of. Rochelle salt into the solution; cobalt and
phosphates also interfere and must be removed. -This
modified method is suitable for the determination of 0·1 to
1·0 mg. of nickel in 100 ml. of solution.
Procedure. Prepare the dimeihylglyoxime reagent by dis-

solving 1 g. of dimethylglyoxime in 100 m!. of rectified spirit.
Prepare the standard nickel solution by dissolving 0·673 g. of
pure nickel ammonium sulphate in water and diluting to 1
litre: ], m!. contains 0·1 mg. of'Ni. The solution may.be
further diluted to a basis of·O·Ol mg. of Ni per ~l., if necessary.
Prepare the nickel solution, containing not more than 1'0
mg. of nickel and free from interfering elements. Render it
slightly ammoniacal, dilute to 100 mI. with water! add 1 m!.
of 1 per cent sodium hypochlorite solution or of saturated
bromine water, followed by 5 mI. of the reagent and mix.
Compare the resulting colour with standards similarly pre-
pared. The colour develops to its maximum intensity in a
few minutes.
AA*
V, 18. Cobalt.-Discussion. a-Nitroso-,8-naphthol (com-
pare Section IV, 33A) may be employed for the determinatioI].
of cobalt for concentratiops between 0·01 and 0·1 per cent;
a red colouration is produc;d with the reagent and 1 part of
cobalt in 1,000,000 parts of solution may be detected.
Iron, aluminium and clrromium are kept in solution by the addi-
tion of ammonium citrate. Copper, nickel and manganese, and. also
preferably iron, should be removed; nickel can be precipitated
with dimethylglyoxime or, better, with the specific reagent a~benzil
dioxime and the excess of the latter destroyed. Nitrates must be
absent. ,8-Nitroso-a-naphthol has been recommended and is said
to be 8 times more sensitive than a-nitroso-,B-naphthol.
Procedu.re. Prepare the a-nitroso-,8-naphthol reagent by
boiling 0·1 g. of a-nitroso-,B-naphthol with 20 ml. of water and
1 m!. of a N s.odium hydroxide solution, filter and dilute to
200 ml. Prepare the standard cobalt ~olution by dissolving
0'4,.77 g. of A.R. cobalt sulphate CoSO,,7H 2 0 in water and
diluting to 1 litre ; 1 ml. contains 0·1 mg. Of Co~ The sollJ.tion
may be dUu'ted 10-fold to a basis of 0·01 mg. of Co per m!. if
required.
Treat the faintly acid (hydrochloric acid) solution (ca.
80 ml.) containing not more than 0·1 mg. of cobalt,.with 5 m!.
of a solution containing ammonium citrate and ammonium
hydrdxide (50 g. of citric acid dissolved in 25 ml.lof water and
50 m!. of concentrated ammonia solution) and dilute-'io 100
ml. Add G m!. of the freshly-prepa.red reagent and mix
thoroughly. Compare the colour produced with those of
standards that have been subjected to the same treatment;
it is best to use the balancing method (Section.V, 4D).
V, 19. Arsenic.-Discussion. Very small amounts of
arsenic (0'001 to 0·1 mg.) are best determined by volatilising
the element as arsine AsH a, and comparing its reaction under
certain experimentaL.conditions with that obtained by the use
of known amounts of arsenic. Various reactions have been
proposed: that due to .Marsh utilises the formation of the
mirror of arsenic produced by the action of heat, w.h.ilst that
due to Gutzeit compares the colourations formed upon discs
of dry paper impregnated with mercuric chloride. A modi-
fication of the original Gutzeit method will be described.
The reduction to arsine is best effected electrolytically in an
apparatus incorporating a mercury cathode or, more simply,
by the older process with arsenic-free zinc ane.! hydrochloric
acid, Both procedures are given below.
An alternative method, due essentially to C. R. Sanger and
O. F. Black (1~07), replaces the disc by a thin strip of mercuric
bromide paper: the arsenic content is given _both by the
depth of the tint and also by the length of the paper bearing
the stain.
The arsenic must b,e in the tervalent state. Nitric acid, the
halogens, and compounds that yield sulphur dioxide, hydrogen
sulphide or phosphine must be absent. The last-named are readily
eliminated by boiling with nitric acid; the excess of the latter is
removed. by evapora.tion with sulphuric acid to fumes of sulphur
trioxide. In such treatment the arsenic becomes quinquevalent
and must be reduced, e.g., with ferrous sulphate solution. The
following metals and their salts are undesirable: silver, mercury,
platinum; palladium, nickel, cobalt and copper. in the presence
of interfering substances, it is usually best to separate the arsenic
first as the trichloride by distillation in an all-glass apparatus from
a solution containing hydrochloric acid and a reducing agent, such
as hydrazine sulphate.
Procedure. Prepare a standard arsenic solution by dissolving
1·00 g..of A.R. arsenious oxide in the minimum volume of
pure sodium hydroxide solution, acidify with dilute hydro-
chloric or sulphuric acid, and make up to 1 litre: 1 m!. con-
tains 1 mg. of·As 2 0 3 • A solution containing 6'001 mg. AS 2 0 3
per m!. is prep'l-red by diluting 1 ml. of this solution to 1 litre.
Prepare the sensitised mercuric chloride papers by immersing
What man arsenic test strips or strips cut from Whatman No.
20 paper in a 5 per cent solution of A.R. mercuric chloride in
95 per cent alcohol, and dry them on a horizontal rack of glass
rods at the laboratory temperature out of exposure to direct
sunlight. Keep them in a desiccator over calcium chloride
in the dark until required.
Prepare the lead acetate papers by soaking filter papers
(10 cm. by 5 cm.)·in a 10 per cent solution of lead acetate in
recently boiled water; dry the papers and keep them in a
closed tube. (These will serve to retain any hydrogen sul-
phide which may be evolved in~the apparatus.)
A. Zinc and acid method. Fit up the apparatus shown in
Fig. 5-17 (a). A is a wide-mouthed bottle of about 100 ml.
capacity fitted with a 2-holed rubber bung. A small tap
funnel B is inserted into one of the holes, and a glass tube C,
20 cm. in length and 5 to 7 mm. internal diameter and drawn
out at the lower end to an aperture of abou~ 2 mm., is passed
through the other hole: a hole of about 2 mm. diameter is
made in C (below the cork) to allow of the free escape of the
gases. The 1l:pper end of C is ground smooth (e.g., on a fine
(a) (b) (0)

FIG. 5-\7.
carborundum stone) at right angles to the tube,. and is fitted

with a device whereby the m.ercuric chloride paper is held
firm1y in position oyer the opening of the -tube, thus ensuring
that all the arsine evolved passes through the paper and a
uniform circular stain is produced. The simplest of these is to
fit a wide tube (about -t"long and ca. 1 "internal diameter) round
the upper part of C by means of a cork and also to provide
another cork, bored with a hole equal in diameter to the external
bore ofthe tube, which fits tightly on to the flat end of C. A
circle of mercuric chloride paper (about 10 mm. in diameter) is
cut out by means of a cork borer and placed on top'of the tube;
the paper is-held firmly in position by inserting the other bored
cork (see Fig. 5-17 (b)). A.]. Lindsey (1930) has developed an
all glass apparatus* (Fig. 5-17 (c)) which serves the same
purpose; the tube is 20 cm.long, with internal bore 5 mm. or
6'5 mm., and is provided with a flange and two ears at the
upper end. The paper is placed on this flange and is held
securely in position by means of a cap with an identical flange,
the upper part of which is worked into a double-lipped shape;
the cap and tube are held together by' means of a pair of
rubber bands.
Roll one of the prepared lead acetate papers in the form of a
cylinder 10 cm. in length and place it so that the upper end is
~
• Supplied by Baird and Tatlock (London) Ltd .• 16-17 Cross Street. London.
E.C.1. England.
"Colorimetric Analysis 717
not less than 2·5 cm. below the upper end of the tube; alter-
natively, the tube may be packed with strips of the lead q.ce-
tate paper about 10 cms. in length.
Prepare a series of standard starns, covering a range of, say,
0·001 to 0·05 mg. of As, as follows. Place 5 to 10 g. of
arsenic-free zinc (e.g., A.R.) in the bottle together with about
0·5 g. of A.R. stannous chloride. Fit the apparatus together,
and place one of the prepared discs of mercuric chloride
paper into position as described above. Introduce 2 ml. of the
arsenic solution (0'.001 mg. per ml.) into the flask, and then
run in' 50 to 55 ml. of 2'5N sulphuric or hydrochloric acid
(prepared from the A.R. acids), and close the tap. Immerse
the bottle in a large beaker-of water containing water at room
temperature. Allow the reaction to proceed for 1 hour, then
remove the mercuric chloride paper disc and dry it in the
dark on a sheet 6f ftlter paper. Repeat the procedure with
fresh discs, using increasing amounts of arsenic solution.
Mark each disc clearly with the exact arsenic content.
If the standards are likely to be used frequently, a fair
degree of permanence may be secured by any of the following
procedures: (i) dry the discs between filter paper, dip them
into cellulose varnish, mount on a sheet of white paper, and
preserve over calcium chloride in a desiccator in the dark;
(ii) dip the (iiscs into molten paraffin wax, and preserve them
over phosphoric oxide in a closed tube which is kept in the
dark; (iii) immerse the discs in a 10 per cent solution of
potassium iodide until the red colour of mercuric iodide at
first produced disappears, leaving a brown stain, and then
dry in the dark on a sheet of filter·paper. Prepared papers
are available commer~ially, which are claimed to preserve
their permanence for 1 year.
If the arsenic content of the solution is approximately
known, the solution is .added to zinc and stannous chloride
cOJ?tained in the bottle, dilute sulphuric or hydrochloric acid'
introdu~ed, and the determination carried out exactly. as in the
preparation of the standards. If the arsenic content is
entirely unknown and interfering substances are absent, a
preliminary test should be carried out as follows. Set up the
apparatus, add the acid to the zinc, and then run in the arsenic
solution in suitable quantities at a time! commencing with 1
ml. and waiting 15 minutes after each addition until a stain is
produced within the range of standards. 1he procedure is the]l
repeated in the manner used in preparing the standards, taking
the appropriate amount of the arsenic solution. The reagents
should first be tested for the presence of arsenic by running a
blank with distilled water.
B. Electrolytic (mercury cathode) method. A mercury
cathode cell, due to T. Callan and Jones (1~30), suitable for the
determination of small quantities of arsenic, is shown in Fig.
5-18 (a).
.. m."' paper
'__'-~u
tubber band
(a) (b)
FIG.. 5-18.
H consists of a flat-bottomed tube, °5" long and 1-3/8* in di~meter,

having an aperture of I" diameter blown on the side aBout i" from
the bottom of the tube. The aperture is covered by a piece of
parchmentised paper,* which acts as an efficient diaphragm of very
'low electrical resistance between the catholyte and anolyte. The
aperture is flanged so that the parchment paper may be held in
position by means of a rubber band. Purified mercury is placed in
the bottom of the tube so that its surface is just below the level of
the aperture. Electrical connexion is made with the mercury 'by
means of a narrow glass tube, into the lower closed end of which is
sealed a short platinum wire. The whole of the exposed platinum
wire dips irfto the mercury; a little mercury is placed insisle the
the tube and a copper wire dipping into this forms the negative
terminal of the cell. The narrow tube, carrying the cathode lead,
passes through a 3-holed rubber stopper which closes the cell. The
.. Parchmentised paper is prepared by rapidly immersing Whatman No: 1
filter paper in pure 80 per cent sulphuric acid for a few seconds, and then
washing with water. The paper is stored under water until required.
Colorimetric A n,alysis 719
rubber stopper also carries a tap funnel of 20 ml. capacity, through
which the electrolyte or test solution may be introduced, and also
an exit tube having a small hole near the bottom and which is con-
nected directly with the p,urifying tube of the Gutzeit apparatus
(not shown). The cathode cell is immersed in the electrolyte of the
anode cell (1 : 8 sulphuric acid), which is itself immersed in a cooling
bath maintained at room temperature. The anode is a sheet of
platinum foil (2" by 1") welded to a short length of thick platinum
wire.
Fig. 5-18 (b) shows the essential parts of the same apparatus, but
with the rubber stopper replaced by a ground-glass joint.*
In use a current of 2 amps. is passed for 15 minutes, followed
by 4 amps. for 15 minutes. In other respects the procedure
for the determination is similar to that given in A.
C. Modified Sanger-Black-Gutzeit method. Prepare the
sensitised mercuric bromide paper as follows. Dry some
Whatman arsenic tests strips No. 130 (120 X 2·5 mm. or
150.X 5 mm.) or strips prepared from Whatman No. 20 paper
in an electric oven at 105°C for 1 hour, and store in a desicca-
tor over calcium chloride until required. Saturate the strips
with a 0·35 per centt solution of pure mer"uric bromide in 95
per cent ethyl alcohol, allow them to drain, and dry in a
desiccator for 10 minutes. The strips are then rcady for use:
no tr~ated strips should be stored for. longer than 2 hours
before use.
Prepare a ferric ammonium sulphate solution by dissolving·
8·4 g. of A.R. iron alum with I ml. of 1 : 4, ·A.R. sulphuric
acid in 100 ml. of water. Ten m!. of this solution contains
the equivalent of,ca. 0·5 g. Fe;Oa.
Prepare a stannous chloride solution by djssolving 16 g. of
A.R: stannous chloride in 20 ml. of water containing 1 ml. of
A.R. hydrochloric acid (sp. gr. 1·2). •
• Both forms of apparatus' are obtainable from Baird and Tatlock (London)
Ltd., Cross Street, London, E.C.I, England.
t Formerly more concentrated solutions (up to 5 per cent) of mercuric
bromide were employed. The weaker the solution, the longer and the less
intense will be the stain. •. •
H. E. Crossley (1936) obtained satisfactory results by the use of .. Swedish ..
filter paper No. IF: strips, without previous drying, were immersed in a 1·5
per cent solution of mercuric bromide in 95 pey cent ethyl alcohol, then
drained for 5 minutes on a rack of glass rods, and finally kept in a desiccator
over calcium c\lloride for 2 hours before use. He utilised strips of 115 x 2,.5
mm. size for As contents of 0·002 to 0·018 mg. and obtained lengths of stains
from 9 to 72 mm.; for As contents of 0·005 to 0·045 mg., he employed strips
of 115 X 5·0 mm and the corresponding lengths of stains were IS to 95 mm.
A 60 ml. generating bottle was used in both cases, but with the higher arsenic
contents the upper tube is of slightly larger bore to accommodate the wider
strip; rea~tion was allowed to take place for 45 minutes,
The apparatus used is shown in Fig. 5-19. It consists of a
60 ml. wide-mouthed bottle into the" mouth-of which is fitted
by means of a rubber stopper a tube
Sirip 0/ 12 em. long and i ·25 cm. in diameter
Merc~riC Bromide Paper
and constricted 5 cm. from the top.
Wal.r Condenser The lower part of the tube contains
a roll or strips of heavy filter paper
soaked in 10 per cent lead acetate
solution and dried (this filter paper
is held in position by a plug of
purified glass wool, which 'also serves
to remove acid spray); the upper
portion contains purified glass wool
soaked in the same solution. This
tube is connected by a rubber stop-
per to a glass tube 10 cm. long and
Glass Wool moistened 4 mm. internal diameter which is
with LBOd Acetal. 80lution constricted 4 cm. from the lower
end; the strip of test paper is
placed in this tupe. In order to
obtain uniform results which are
independent of the temperature of
the laboratory, a small water con-
denser, through which water below
20°C is circulated, surrounds the
Glass Wool
tube containing the test-strip.
Place the solution under test
(containing not more than 0·02 mg.
_80lullon of As 20 a) in the generating bottle,
add sufficient acid so that the acid
content is equiValent to 4·2 to 6·3 g.
of sulphuric acid, then introduce 2
ml. of the ferric solution* and
FIG. 5-19.
0·5 ml. of the stannous chloride
solution: dilute to 50 m!. Intro-
duce about 15 g. of arsenic-free zinc (preferably in the
form of shot, 1/3 to 1/6" mesh). Immediately insert the
connecting tubes, an<i place the mercuric bromide strip (120 X
2·5 mm.) in position; shake the bottle gently and allow to
stand for 60 minutes. The temperature should be main-
tained at 20° ± 1°C (or, less satisfactorily, at 25° ± PC) by
• Ferrbus iron prevents polarisation between the zinc ancl acid, and llence
I1ccelerates t]l.e liberation of the arsine,
immersing the bottle in a water bath or a large beaker con-
tainihg water at this temperature. Remove the strip, dry it
between sheets of filter paperl dip it into either cellulose
varnish or into molten paraffin wax (free from water), and
mount it on a sheet of white paper. Compare the stain
produced with stains obtained in the same manner by known
amounts of arsenic, using portions (,f a standard solution
containing 0·001, 0'002, 0'004:, 0'006, 0'01, 0'015, and 0·02
mg. of As 20 a and adding water, acid, etc., so as to have the
same volume and concentration as the final unknown solution.
The strips bearing the standard stains should be mounted on a
sheet of white paper, folded back to form a cylinder and
preserved in a closed test tube containing phosphoric oxide
in the dark. A blank determination must of course be carried
out with the reagents; this should show not more ~ha'It ~·001
mg. of As 2 0 a. Instead of direct comparison with standards,
the lengths of the stains by a given procedure may be plotted
on a graph and values interpolated; results accurate to 5
per cent are obtained.
For the determination of quantities of arsenic varying
'between 0'02 mg. and ·0·5 mg. of AS 20 3 , a larger apparatus is
employed, The bottle has a capacity 6f 250 mI., the lower
tube is 15 cm. long and of 1·5 cm. diameter with a constriction
6 em. from the upper end, whilst the upper tube is 18 em. long
and of 7·5 mm. bore and is constricted 9 cm. from the lower
end. The strips of paper should be 150 X 5 mm. and are
similarly prepared with a 3·5 per cent solution in 95 per cent
ethyl alcohol. The final volume of the solution in the bottle
should be 200 m!. and should contain an equivalent of 18·5 g.
of sulphuric acid and 0·1 g. of ferric oxi<\e and also 0·5 mI. of
the stannous chloride solution; about 40 g. of A.R. zinc is
required and reaction is allowed to continue for 40 minutes.
The standard stains should be prepared with 0·02, 0·05, 0,1,
0·2, 0·3, 0·4 and 0·5 mg. of AS 20 3 • The quantity of the
unknown solution employed should be such a~ to contain
between 0·02 and 0·5 mg. of As 2 0 S'
V, 20. AntiinollY.-Discussion. The most satisfactory
method for the colorimetric determination of small amounts
(up to 0·5 mg.) of antimony is the potassium iodide-pyridine
method (S. G. Clarke, 1928), and is particularly well adapted
to the determination of antimony in tin. The preliminary
separation is effected by deposition on copper foil (Reinsch's
reaction). The deposit is dissolved in sodinm peroxide sol11-
tion, and any dissolved copper is precipitated by hydrogen sul-
phide in the alkaline solution and the copper sulphide removed
by filtration. The solution. is acidified with sulphuric acid
and evaporated to fuming; the' residue is diluted with water
and to it are added aque_ous solutions of pyridine, sulphur
dioxide, potassium iodide, and gum arabic or " starch glycer-
ite."'" The yellow colouration is then compared against
standards similarly prepared.
Tin and arsenic when present in moderate amounts (several
centigrams) do not interfere. Bi~muth and several other heavy
metals give a coloured precipitate, whilst zinc yields a white crys-
talline precipitate: these metals should therefore be absent. :r'he
best acid to employ is sulphuric acid. Hydrochloric acid and
particularly chlorides bleach the colour. Excessive amounts of
sulphur.ous acid cause liberation of iodine, whilst a large excess of
pyridine produces colour fading.
Procedure. Full details for the determination of small
amourits of antimony (0,001 to 0·05 per cent) .in tin will be
given.
Prepare. a standard solution of antimony by dissolving
0·2764 g. of A.R. t-e.rtar emetic in 1000 m!. of 10 per cent
sulphuric acid; 1 ml. contains 0·1 mg. of Sb. .
W!'!igh' out accurately a suitable quantity of tin (1 to 5 g.
according to the antimony content, which shol_lld be ca.
0·5 mg.) into a 75Q ml. conical flask; add 50 ml. of concen-
trated hydrochloric acid and sufficient bromine to ensure an
excess while the tin is dissolving. When the tin has dissolved,
add 2 to 10 g. of oxalic acid (accordjng to. the amount .of
antimony), followed by about- 350 mI.. of water. Boil the
liquid gently: the cfxalic acid will dissolve and yield a solution
tinged with bromine. Add about 0·5 g. of sodium hypo-
pjlOsphite to reduce all the free bromine, to separate com-
pletely any arsenic Which may be present, and to render the
solution colourless. Introduce a flat open spiral of copper
(made from a strip of clean copper foil, 15 X 2 em.) into the
solution; it should stand upright on the bottom of the flask.
Boil for 2 hours, remove the copper coil by mean? of a hooked
glass rod, and wash it successively with a gentle stream of 5
per cent hydrochloric acid and of water. Place the coil in a
• This is prepared by heating starch with glycerine and diluting with water.
This (and also gum arabic) is added as a protective colloid to stabilise the
suspension of the yellow complex. It also gives a bluish-green colouration
with free iodine, and then will indicate ~hether tqc solution is cO!llpletely
reduce<! by the sulphur dioxide added.
Colorimetric A natysis 723
small beaker, cover it with water, add 1 to 2 g. of pure sodium
peroxide, an<;l warm until the deposit _js dissolved and the
'copper is well darkened by oxidation. Pour off the solution
from the coil" and then rinse it with water. Pass hydrogen
sulphide into the solution and washings for 15 seconds, allow
the small precipifate of copper sulphide to coavulate on the
water bath (15 minutes) and filter; wash with a dilute solu-
tion ,of ammonium nitrate. Add 5 ml. of concentrated
'sulphuric acid and a few drops of concentrated nitric acid to
the filtrate and washings, evaporate just to fuming, cool and
add 15 ml. of water. The antimo~y in the resultant solution
(ca. 20 ml.) is determined colorimetrically as follows.
Introduce the re!1gents in the order given iato a 100 ml.
Nessler cylinder: 10 ml. of 1 per cent gum arabic solution, 5
ml. of 20' per cent potassium iodide solution, 1 ml. of 10 per
cent aqueous pyridine, 1 ml. of sulphur dioxide solution pre-
pared by lO-fold dilution of.a saturated aqueQus solution, and
60 ml. of cold 1 : 3 sulphuric acid. The antimony solution
prepared as above is then added, the beaker being .rinsed
with not more than 5 ml. of water, and the whole stirred with
the special glass stirrer (see Section V, 4B). Run the stana.ard
antimony solution from a burette into another Nessler tube
containing similar quantities of the reagents (except that 80
ml. of 1 : 3 sulphuric acid are used instead of 60 mI.) until the
colours match after the solution has been well stirred. The
final adjustment is made by adding a small quantity of water
to make the volumes in the two Nessler tubes equal. Not
more than 10 ml. of the standard antimony solution should be
required for otherwise the colour will be too deep for accurate
comparison. A blank determination should be made on the
reagents. .
V,21. Tin.-Discussion. A dilute solution of stannous tin
(concentration::} 30 mg. per litre) gives a blue colouration
with an acid molybdate solution. The composition of the
"molybdenum blue complex" appears to be fixed if the
concentration of the tin does not exceed 30 mg. litre; this
forms the basis of a method for the'colorimetric determination
of tin. The solution should be cold, nearly neutral and free
from excessive amounts of neutral salts; a trace of phosphoric
or arsenic' acid makes the complex more stable. Antimony,
and oxidising and reducing agents must be absent.
Procedure. Prepare a standard solution of tin by dissolving
1·00 g. of A.R. tin in 100 ml. of 1 : 1 hydrochloric acid and
724: Quantitative Inorganic Analysis
dilute to I litre with the same acid; 1 mI. contains 1 mg. 9f
Sn. For the preparq.tion of more dilute solutions, dilution
should be effected with I : I hydrochloric acid. Prepare the
molybdate reagent by dissolving 1 g. of A.R. molybdenum
trioxide in a little sodium hydroxide solution, acidify the
solution with hydrochloric acid, and dilute to 200 ml.
Treat the acidified solution containing up to 1 mg. of tin
with a little pure aluminium, and boil in an atmosph~re of
carbon dioxide to reduce the element completely to the stan-
nous condition. Take a measured volume of the solutio.n,
add 2 ml. of the molybdate reagent and, after 1 minute, 10
mL of amyl alcohol; this will extract the whole of the blue
coloured compound. Separate the amyl alcohol layer,
clarify-if necessary-by the addition of 1 mL of ethyl
alcohol, and compare with standards similarly prepared.
V, 22. Lead.-Discussion. For the determination qf
minute quantities of lead (0'005 to 0·25 mg.) advantage is
taken of the fact that when a sulphide is added to a solution
containing lead a brown colour, due to the formation of
colloidal lead sulphide, is produced.
Interference is produced by the presence of (a) neutral salts, such
as ammonium chloride and particularly tartrates and citrates,
and (b) other metals, such as copper, bismuth, iron and aluminium.
Errors due to (a) may be allowed for by ensuring that the standards
for comparison contam amounts of salts approximately equal to
those in the solution under test, whilst those produced by (b) are
eliminated by the usual analytical procedure or by the use of diphenyl-
thiocarbazone (see belbw). The disturbing effect due to copper
and iron if present in small amount may be overcome by the addition
of a few drops of a 10 per cent aqueous solution of potassium
cyanide; aluminium may be retained in ammoniacal solution by the
addition of ammonium citrate solution, a corresponding amount of
the latter being added to the standards.
In case of doubt as to the presence of other metals it is
best to isolate the lead quantitatively by means of diphenyl-
tbiocarbazone CeHs.NH.NH.CS.N = NC6HS ("dithizone "). The
reagent consists of a soilltion in carbon tetrachloride or,
preferably, chloroform in which it is more soluble, and yields
a red colouration with minute quantities of lead salts. A
number of metals interfere, but the reaction of most of these,
with the exception of bismuth, is inhibited by the addition of
aqueous potassium cyanide solution. The lead complex is
removed by extraction with chloroform, but the resultant
coloured lead compound is not well adapted for colorimetric
determination; the organic lead complex is oxidised and ~he
lead is ultimately determined by the colloidal sulphide method.
Procedure. Prepare. a standard lead solution by dissolving
either 0:183 g. of A.R.lead acetate crystals or 6,160 g. of A.R.
lead nitrate in 100 ml. of'wate.r; 10 m!. of this are diluted to
100 ml. for a working solution, of which 1 m!. contains 0·1 mg.
of Pb.
As an illustration of the simple lead sulphide procedure, the
determination of lead in commercial tartaric acid will be
describe'd. Dissolve 10 g. of the tartaric acid in about 40 m!.
of hot water, add 5 to 6 drops of 10 per cent potassium cyanide
solution, then 25 mI. of 1 : 2 ammonia solution (whi~h should
make the solution ammoniacal), and finally 2 to 3 drops of
ammonium sulphide solution (prepared by saturating 1 : 4
ammonia solution with well-washed hydrogen sulphide).
Make up to 100 m!. Match the solution thus prepared
against a standard solution made by treating 10 g. of lead-
free tartaric acid in exactly the same way and adding to it 2
ml. of the standard lead solution. If the acid being tested
yields an appreciably darker colour than the standard, a
smaller quantity of it should be taken and subJ'ected to
a simtlar treatment. A darker standard should not be
used.
The following is. a general procedure and is based upon the
use of dithizone. If organic matter is present, destroy it by
digestion with sulphuric and nitric acids, dilute, and evaporate
to fuming to remove the nitric acid. Cool the residue, dilute
and treat successively with 2 g. of citric acid, ammonia solution
until alkaline, and 1 ml. of 10 per cent potassium cyanide
solution. The solution should n0W have a volume of 100 to
150 mI. Extract it in a small separating funnel with three
successive portions (10 ml., 5 ml., 5 ml.) of ~ freshly prepared
solution of 0·1 g. of diphenylthiocarbazone in 100' mI. of
chloroform; continue to extract with small portions of
solution (say, 2 mI.) until the extract no longer has a red
colour. Wash each extract in succession with 10 m!. of water,
combine the extracts, and distil off the solvent. De9troy the
organic part of the residue by heating wjth 0·5 ml. of concen-
trated sulphuric acid and a few drops of nitric acid. Di!?solve
the residue in water, add 2 g. of A.R. ammonium acetate,
ammonia solution until alkaline, 1 m!. of 10 per cent potassium
cyanide solution, and 0·1 ml .. of 10 per cent sodium sulphide
solution, Compare the colour developed with standards
similarly prepared (and also containing 2 g. of ammonium
acetate) from the standard lead solution.
V, 23. Copper.-Discussion. Potassium ferro cyanide reacts
with very dilu~ neutral solutions of copper salts to. yield a
purplish-brown colouration due to oopper ferrocyanide. In
the presence of ammonium nifrate, the reaction will serve
to detect 1 part of copper in 2,500,000 parts of water, and
m:3:Y be used as a basis for the colorimetric determination of
small quantities of copper (up to 0·5 mg.).
The solution should be neutral, for fref> acid lessens the colbur and
changes it from a purplish-brown to an earthy-brown. If acid, it is
rendered slightly alkaline with ammonia solution and the excess of
ammonia boiled off; if alkaline, because of the presence of caustic
alkalis, the solution is barely acidified by an acid and then treated
with a little ammonia solution, etc. Lead, when present in too large
a quantity, has little effect upon the colour. Iron interferes, how-
ever, but can be removed as follows. The solution is treated with a
few drops of <concentrated nitric acid to oxidise the iron to the ferric
state, and the iron is subjected to a\ double precipitation with
ammonia solutio}} in the presence of 1 g. of ammonium sulphate.
The filtrate' and washings ru:e boiled to remove ammonia c0mpletely
and the copper determined in the residual solution.
A more sensitive reagent is sodium diethyldithiocarDamate
(C2H6)2N.CS.SNa (T. Callan and J. A. R. Henderson, 1929),
which yields a golden-brown colouration with minute amounts
of copper salts. It will detect 1 part of copper in 50,000,000
parts of solution. The colour is produced in either neutral,
acid or ammcmiacal solution.
The reaction is not specific for copper: a number of other metals,
when present in appreciable quantity, give a white turbidity or
colouration. Thus silver, lead, mercury, cadmium, antimony, tin,
aluminium, chromium and zinc yield white turbidities, whilst
bismuth, iron, manganese, nickel and cobalt give rise to coloured
compounds. Lead may be removed as sulphate. Separation from
iron may be effected either by double precipitation with ammonia
solution as in the ferrocyanide method given above or by precipita-
tion of the copper as sulphide.at pH 3. The best procedure, how-
ever, appears to be to extract the coloured copper complex with
carbon tetrachloride: iron is retained in solution by ammonia and
ammonium citrate solutions, but bismuth must be removed first.
Procedure. Prepare a standard copper solution by dissolving
0·3928 g. of A.R. crystallised copper sulphate in 1 litre of
water; 1 mI. contains 0·1 mg .. "Of copper. For the dithio~
carbamate method, it is best to dilute this solution lO-fold.
A. Ferrocyanide method. The duplication method (Sec-
tion V, 4B) is often employed for this determination. Secure
two 100 ml. Nessler cylinders and place 5 elrops of a 4 per cent
aqueous A..R. potassium ferrocyanide solution in each. Add
a measured quantity of the neutral solution (free from inter-
fering element~) in which copper is to be aetermined to one of
them; fill both to the mark with distilled water. Add 5 ml.
of 10 per cent A.R. ammonium nitrate solution to each, and
then the standard copper solution from a burette to the blank
until the colourS"in both cylinders match; stir the liquid well
after each addition. The exact volume of the standard
copper solution added is note4 and from this the amount of
copper in the unknown solution can be calculated. For
further general details, see Section V, 4B.
B. Sodium diethyldithiocarbamate method. The copper
solution should be free from interfe.ring elements and should
contain not more than 0·05 to 0·08 mg. of copper; it should
also be either ammoniacal or acid with nitric acid. Place a
known volume of the copper solution in a Nes~ler cylinder, add
10 m!. of a 0·1 per cent aqueous solution of the reagent and
make up to 100 ml. Compare the colour produced by the
sta1J.dard series method (Sectiop V, 4A) with that of standards
prepared in exactly the same way. J'he addition of a, little
protective colloid, e.g., gelatin, fo the solutions increases their
brilliance, and facilitates accurate comparison of the colours.
If the e~traction method is to be employed, proceed in the
following manner. If iron is present, convert it into the ferric
state, a,dd 2 g. of citric acid and render alkaline ':Vith ammonia
solution. Then add 10' m!. of a 0·1 per cent aqueous solution
of the reagent, and extract four times with three 5 m!. portfons
of pure carbon tetrachloride-the final extract should be
colourless. Compare the colour produced with tl)at of solu-
tions containing known quantities of copper treated in a
similar manner. ..
V, 24. Bismuth.-Discussion. When potassium iodide
solution is·added to a dilute sulphuric or nitric acid solution
containing a small amount of bismuth, a yellow' colouration,
due to the forma.tion of a complex iodide, is produced. This
is suitable for the colorimetric determination of quantities
up to about 1 mg. ,of bismuth.
The solution must be free from large amouhts of lead, copper, tin,
arsenic, antimony, gold and silver, and from elements in sufficient
quantity to colour the solution, e.g., nickeL Substances which
liberate iodine from potassium iodide interfere, for example, ferric
iron; the latter should be reduced by sulphur dioxide and the excess
of gas boiled off. Separation of bismuth from copper is best effected
by making use of the insolubility of bismuth phosphate in dilute
nitric acid.
For the determination of bismuth in lead, the latter is
precipitated as sulphate in the presence of tartaric acid (to
prevent coprecipitation' of the bismuth), the tartaric aCid is
destroyed· by wet oxidation with nitric and sulphuric acids,
the nitric acid eliminated by evaporation· to fuming, the
solution diluted, and sulphurous acid and potassium iodide
solution' added. The yellow colour is then compared w:th
standards similarly prepared.
Procedure. Prepare' a standard solution of bismuth by dis-
solving 0'105 g. of A.R. bismuth* in 20 m!. of concentrated
sulphuric acid, and dilut.e to 1 litre with water; 1 m!. con-
tains 0·1 mg. of Bi.
Treat the colourless solution (20 to 30 mI.), containing not
more than 0·5 to 1 mg. of bismuth and free from interfering
substances but acid with sulphuric acid, with 1 mI. of a
saturated solution of sulphur dioxide, 10 m}. of a 30 per cent
solution of potassium iodide, ~nd dilute to 50 ml. Matchihe
yellow colour produc~d against standards containing the
same concentration of sulphurk and sulphurous acids. After
the determination add starch 'Solution to make sure that the
colours are not partially due to free .iodine. '.
V, 25. Silver.-Discussion.-P-Dimethylamino-benzal-
rhodanine (CH 3 )2N.C6 H 4 .CH= C.CO.NH.CS.S gives a pink to
'---:--:----:---::-'
red colouration with small quantities of silver, and may be
used for its colorImetric determination in solutions containing
0·06 to 9 mg. per litre. Soft soda glass adsorbs silver, and it
is best to e,mploy silica apparatus; Pyrex glass-ware may be
used for determinations at l=oom temperature, but cannot be
employed for the evaporation of silver-containing solutions.
The method is particularly useful for the determinatioJl of
silver in drjnking water, Swimming pools, or those' cases
where silver has been employed as a sterilising agent.
Cuprous copper, mercurous mercury, gold, platinum and palla-
dium interfere. Calcium carbonate, sodium nitrate and ammonium
nitrate increase the depth of colour due to silver if any of them are
present in concentrations greater than 1 mg. per litre. Very small
• Obtainable from the Mallinckrodt Chemical Works, New York, U.S.A.;
see, however, Appendix, Section A. 9.
amounts of either sulphuric acid or Qf potassium sulphate (0'6 g.
per litre for the latter) also increase the depth of the rhodanine
colour. Chlorides, even when present in small amounts, cause a
fading of the colour; they are removed by 2 or 3 evaporations of the
solution in a silica vessel with concentrated nitric acid.
Procedure. Prepare a series of suitable standards from A.R.
silver nitrate. Treat the neutral solution, containing not
more than 9 mg. of silver per litre and free from the interfeting
substances mentioned above, with 0·5 tnl. of 4N nitric acid-
and sufficient water to make the volume 15 ml. Mix thor-
oughly, add 0·5 ml. of a 0''02 per cent alcoholic solution of
p-dimethylamino-benzal-rhodanine, ana. mix again to develop
the pink or red colour. Match the colour against standards
similarly prepared from silver nitrate soluti_on.
V~ 26. Silicate.-Disc~ssion. Small quintities of dissolved
silicic acid react with a solution of a molybdate in a mineral
acid medium to give an intense yellow colour, due probably to
... the cQmplex acjd Hs(Si(Moa07)u]' The colour is similar to that
of potassium chromate solutions, which can therefore be used
for the preparation of- perman~nt standards for the colori-
metric determination of silicates. The metho~ is applicable
up to about 50 mg. of ~_ilica per litre.
Phosphates and arsenates give similar colourations and should
best be absent; it is said, however, that small amounts of phosphates
do not interfere, but it is best to remove them by magnesia mixture
or render them inactive by the addition of ammonium citrate.
Reducing substances, such as ferrous iron and hydrogen sulphide
give a blue colour, and must therefore be rerl\.oved. Iron in greater
concentrafion than 20 mg. per litre interferes. Water used for
dilution must be freshly distilled, for it dissolves significant traces
of silica on standing in glass.
The method is chiefly employed for the determination of
silica in water supplies and natural waters, and its use in this
connexion will be described.
Procedure. Prepare the standard comparison solution by
dissolving 0·630 g. of A.R. potassium chromate in 100 ml. of
water; 1 ml. of this solution is equivalent to 1 mg. of SiOI'
Prepare the reagent by dissolving 30 g. of A.R. ammonium
molybdate in 200 ml. of 1 : 1 hydrochloric acid and 400 ml.
of water.
Place 50 ml. of the sample of water into a Nessler tube, and
add 5 ml. of the molybdate reagent. Stir thoroughly, and
after 10 to 15 minutes compare the yellow colour developed
with the chromate standard as follows. To a similar Nessler
cylinder c.ontaining 50 ml. of water* add .the standard chro-
mate solution (x ml.) from a burette until the colour matches
that of the unknown solution. Repeat the determination by
pladhg (50-x) ml.,pf water '(or of 0·5 per cent borax solution)
in the Nessler cylinder, and add the chromate solutIon until
the colour matches again. One ml. of the standard chromate
s~u'tion corresponds to 1 mg. of SiO z per 50 ml. of the sample
or 20 mg. per litre of sample.
V, 27. Detennination of othet elements and radicals.-
Space does not permit the inclusion of a detailed acyount of the
determination of small quantities of other elements and
radicals. The following Table refers to selected methods
which are availa15le for elements and radicals not alre;:tdy
discussed, and reference is made to standard works where full
details will be found.
1 = E. D. Snell and C. R. Snell, Colorimetric Methods 01.
Analysis, Vol. I (1936).
2 = N. H. Furman and W. W. Scott, Standard Methods of
Chemical Analysis, Vols. I and II (1939).
3 = J. H. Yoe, Photometric Chemical Analysis, Vols. I and
II (1928). ~'
Element or
Radical Method or Reagent Detail* in
-
Mercury (a) Mercuric sulphide I, 3
(b) Diphenyl-caxbazone 1
Thallium (a) Thallous sul'phide 1
(b) Thallous phosphomolybdate 1 r
Gold (a) Prister's method 1,2,3
(b) Cassel's method 1,2,3
(c) Moir's-method 1,2, 3
Platinum (a) Complex iodide 1,3
Beryllium I (a) Quinalizllrin 1
(1 : 2: 5 : 8-tetrahydroxy-anthra-
quinone)
Cerium (a) Gallic acid 1
Zinc (a) Zinc sulphide 1
(b) Zinc ferrocyanide 1,3
Calciupl (a) Alizarin 1,2
(b) Picrolonic acid 1
Magnesium (a) .. Magneson " (para-nitrobenzene-azo- 1
resorcinol)
(b) Curcumin 1
Sodium (a) Sodium uranyl zinc acetate
I I
.
• It is better to use 50 m!. of 0'5 per cent A.R. borax solution, as the colOUr
is then more stable and more easily repl:oducible.
Element or
Af~thod or Reagent Details In
Radical
Potassium (a) Potassium chloroplatinate and reduction

with stannous chloride 1, 3
(b) Potassium 'Sodium cobaltinitrite 1
Chloride (a) Silver chloride I
Chlorine (a) o-Tolidine 1,2
(b) Benzidine 1
Chloramine (a) o-Tolidine 1
(h) Nessler's reagent 1
Bromide (a) Schiff's reagent 1,2
(b) Free bromine and carbon tetrachloride 1,3
(e) Fluorescein 1
Iodide (a).,palladous chloride 1
(b) Starch-iodine reaction 1
(e) Bromine water 1
Fluoride (a) Bleaching effect upon oxidised titanium
solution 1, 2,3
(b) Zirconium oxychloride and purpurin
(1 : 2: 4-trihydroxy-anthaquinone) 1,2
Cyanide (a) Ferric thiocyanate 1,3
(b) Prussian blue 1,3
(c) As ammonia"by Nessler's reagent 1
Oxygen (a) Ammoniacal cuprous chloride 1,2,3
(b) Starch-iodine complex 1
(e) Indigo-carmine 1
Hydrogen (a) Oxidation of ferrous iron, followed by
peroxide thiocyanate reaction 1,3
(b) Ammonium molybdate 1,3
Ozone (a) Fluorescein 1
(b) Destruction of nitrite I
Nitrate (a) Sodium diphenylamine sulphonate 1,2
(b) As ammonia by Nessler's reagent 1, 3
(e) Phenol disulphonic acid 'I, 3
Phosphate (a) Phosphomolybdate, reduction with stan-
nous chloride 1
(b) Phosphomolybdate, reduction with
hydroquinone 1,3
Borate (a) Turmeric paper 1,2,3
(b) Curcumin 1,3
Sulphide (a) Lead sulphide 1,2,3
(b) p-Amino-dimethyl-aniline
• (methylene blue method) 1,3
Sulphate (a) Barium sulphate 1
(b) Benzidine 1
CHAPTER VI
GAS ANALVSIS
VI, 1. General discussion.-The methods of gas analysis
may be classified as follows :-
1. Volumetric methods.* A measured volume of the gas
mixture, at a known temperature and pressure, is subjected
to selective chemical r,eactions (absorbents) to remove the
various constituents in succession, the quantjties so removed
being determined by the resultant diminution in volume.
When there is no suitable absorbent for the gas to be deter-
mined (e.g., methane), it is frequently possible to mix it with
excess of another gas (e.g., oxygen) with which it reacts
chemically (often in the presence of a catalyst) and under
proper conditions to deten;nine t~e change in volume which
takes place.
In a precision form of apparatus, due to Bone and his co-
workers (see Section VI, 6E), the gas is expanded to a fixed
volume and the pressure exerted is measured, as is also the
diminution in pressure of the gas after each absorption
when expanded to the same fixed volpme. The constant
-volume method may also be employed with the Ambler
portable apparatus (Section VI, 6D).
2. Titration methods. The constituent sought is absorbed
in a liquid reagent of knoWn concentration. This method is
frequently employed for gases, such as sulphur dioxide, which
are-soluble in water. An interesting example is the deter-
mination of small quantities of carbon monoxide by passage
over heated iodine pentoxide : •
12 0 6 + 5CO = 5C0 2 + 12 ;
the liberated iodine is titrated with dilute standard sodium
thiosulphate solution.
3. Grawnetric methods. The gas is absorbed by, or reacts
with, a weighed amount of the reagent, or else is absorbed in
a liquid aud a precipitate is subsequently weighed. Thus
sulphur is determined in coal gas (Referees' method) by
burning a known volume in a special Bunsen burner; the ~ir
for the combustion passes over a shelf containing ammonium
• The term volumetric in this connexion is restricted to 1Cesses in whic\l
gas volumes are measured. I
732
Gas Analysis 733
sesquicarbonate and the ltberated, ammonia' carries the sul-
phur dioxide with it. The gases are absorbed in water, the
resultant solution treated with a little pure bromine, acidified
with dilute hydrochloric acid, warmed to remove excess of
bromine, and .the sulpbur determined by precipitation as
barium sulphate in the usual manner.
4. Physical methods. The constituents are determined by
measurement of their physical properties, such .as density,
thermal conductivity .. viscosity and refractivity.
The following account of gas analysis is intended merely
as an introduction to the subject, and will be considered under
two headings: (a) analysis of gaseous mixtures, and (b) the
determination of the volume of a gas produced in a chemical
reaction and its application to the analysis of solids and
liquias. For further details, the reader is referred to th~
standard text books on gas analysis (see Appendix, Section
A,3).
VI, 2. Sampling.-The first step in the analysis of a gaseous
mixture is the collection of a representative sample of the gas
FIG. 6-2.
(a)
FIG. 6-1.
m a suitable vessel. If a large quantity of the gas is available

it may be either passed or aspirated through a sample tube
of the type shown in Fig. 6-1 (a) or (b). The former has two
2-way capillary taps: the latter has two 3-way capillary stop-
cocks, which pePffiit the flushing out of the connecting tubes
with the sample itself or with a confining liquid. When the
quantity of gas available- is limited, the apparatus shown in
Fig. 6-2 may be used: the connexions are sleared of air by
blowing or aspirating some of the :gas through the upper tap,
or alternatively by evacuating the connexions from the same
place and filling them with m~rcury, The gas is collected in
S by lowering the reservoir R.
If a sample of the gas is to be collected and kept for analysis
for a considerable time, the por-
tions of gas taken f~r analysis
must be displaced with mercury,
A small gasometer that is suit-
able for prolonged storage is
.shown diagrammatically in Fig,
6-3. The gas is contained in
the large glass bulb A, which
carries a capillary U-tube B at
the top, and at the bottom is
connected to the levelling bulb
~ by means of a rubber tube·;
the capillary is closed by means
of a rubber tube and pinch-cock
D. The apparatus is first filled
with mercury: by lowering or
FIG. 6-3. raising the levelling bulb, gas can
. be drawn in or driven out as desired. If the gas sample is to
be kept for some time, the capillary 'tube B is filled with
mercury by inserting q. small pipette at D. A simple gaso-
meter may also be constructed from a round-bottomed flask
fitted with a 2-holed rubber
stopper and bent glass tubes
as shown in Fig. 6-4: the
right-hand tube is connected
to a levelling bulb by means
of rubber tubing as in Fig.
6-3.
VI, 3. Purification of mer-
cury.-The mercury em-
ployed in analysis should 'be
carefully purifie.d, If im-
pure, it clings to the glass,
Gas Analysis 735
and accurate readings are impossible. The principal Im-
purities in ordinary mercury are copper, cadmium and
zinc. The impurities are largely removed
by allowing the mercury to run in a fine
stream through a long -length of 5 per
cent nitric acid in the apparatus shown
in Fig. 6-5. A glass tube of about 1 inch
Dore and 20 to 30 inches length is fused to
an S-shaped glass capillary of about 2
mm. bore. The main tube is filled with
5 per cent nitric acid, and the siphon is
so shaped that the bottom of the tube
always contains a little mercnry. An
ordinary funnel cO!ltaining a filter paper
folded in rl].e usual way is supported
above the apparatus with the stem just
dipping into the nitric acid. A small
• hole is then mad~ in the apex of the filter
paper by means of a pin, and the mercury
poured into the filter. The mercury
usually takes a zig-zag course as it flows
in the form of a fine spray through the
dilute acid; any dust, etc., contaminating
the mercury is retained in the filter,
together with a little mercury. The mer- FIG. 6-5.
cury collects in the bottom of the tube,
and siphons over into the collecting bottle or beaker. The
~ proce!iS !>hould be repeated a second time, and the mercury
finally passed through a column of water in a similar apparatus.
The mercury may be dried by heating to 100 0 to 150 0 C
(preferably under reduced pressure) in a fame cupboard
provided with an efficient draught. MERCURY VMOUR IS
mQRLY POISONOUS AND THE EFFECT IS CUMULATIVE;
hence exposure of mercury to the open atmosphere should be
reduced to the absolute minimum and all operations with mer-
. cury at elevated temperatures must be conducted in a fume
cupboard. All bottles containing mercury should be stoppered.
For most 'purposes 'the mercury obtained by the above
process is suitable for use. If mercury is to be used for
calibrating measuring apparatus, it is best distilled under
reduced pressure in a slow stream of air; under these condi-
tions the vplatile metals, cadmium, zinc, etc., form non-
volatile oxides, and complete purification of the mercury from
these metals can be obtained.
All apparatus utilising mercury should be set up on a
wooden base having a raised edge, which will s~rve to retain
any mercury spilt during any of the operations.
The base should be provided with a small hole,
ordinarily closed with a cork, through which the
mercury may be swept and recovered. For picking
up spilled globules of mercury, the special pipette
shown in Fig. 6-6 is useful ~ suction may be applied
by filter pump or mouth through a rubber tube.
The bulb forms a reservoir or safety·trap.
VI, 4. Correction of volume of gas for tem~
. perature and pressure.-The volume of a gas dep.ends
upon the temperature and pres-sure to which it is
subjected. In all cases where a gaseous volume is
measured the observed volume must be reduced
FIG. 6-6 to the corresponding value under normal conditions
of temperature and pressure. The most generally employed •
• temperature is O°C, and the most generally employed pressure
is that due to a column of mercury 76·0 em. in height and of
density 13·596 (i.e., at O°C): these are usually termed N.T.P.
The variation of the density of mercury over the range O°C
to 30°C is 0,0738, thus an error of about 0·5 per cent would be
introduced by a variation of 30°C of the temperature of the
barometer, Table XXV contains the values of the density
of mercury at various temperatures.
TABLE XXV. DENSITY OF MERCURY (IN GRAMS PER ML,)
Temp.
DC. 0 2 4 6 8 110 12
-----
I 14 16 18
_20° 13·645 13·640 13'635 13'630 13 '625 13 '620 13'61513·610 13·605 13·600
-- - - - - - -- - - - - - - - - - - -- -
0° 13·596 13·591 13·586 13·581 13·576 13·571 13·566 13·561 13'556 13·551
-20°
-- - - - - - - - - -- -- -- - - -- -
- -- - _-
13·546 13'541 13·536 13·532 13'527 13·522 13·517 1,3'512 13·507 13'502
- - - - - - -- - - ~
- -- -
--
---
40° 13·497 13·492 13·488 13·483 13·478 13'473 13-468 13·463 13·458 13·453
- - - - -- - ---- - - - ---
60° 13·449 13·444 13·439 13·434 13·429 13·424 13'420 13'415 13'410 13'405
- - - - - - - -i----' - - - -
80° 13·400 13·395 13·390 13·386 13·381 13'376 13'371 13·366 13·362 13,357
- -- -- - - -
100° 13·352
I
The forl!).ula for applying the correction is :
Ho = HI {I - (fJ - a}t}r
:r
Gas Analysis 737
\
where HI is the height of the barometer ,at tOe, Ho is the
reduced height at oDe, f3 is the coefficient of linear expansion
of mercury, and, a is the coefficient of expansion of the
material of the barometer scale.
If'V is the volume of dry gas measured at,a pressure of P
~m. of mercury (corrected for tempe!ature as already ex-
plained) and a temperature of tOe, then it follows by application
or Boyle's and Charles' laws that the volume Vo at N.T.P.
is given by:
273.
(273 +t)
If the gas is measured wet, i.e., saturated with water vapour,
the pressure P must be reduced by the vapour pressure of
water at the temperature of the measurement. Table XXVI
contains the values of the vapour pressure of water at 'various
temperatures.
TABLE XXVI. PRESSURE OF AQUEOUS VAPOUR (IN MM. OF MERCURY),
Temp. Pressure. Temp.

.
Pressure Temp. Pressute.
°C °C. . ac.
0' 4-60 10' 9.17 20° 17'39
1° 4'94 11° 9·79 21 0 18'SO
2° 5'30 12° 10'46 22° 19·66
. 3°
4°
5°
5'69
6'10
6:ro
13°
14'
ISO
11-16
11·91
12·70
23?
24'
25'
20·S9
22·18
23'55
6° 7'OQ 16' 13'54 26° 24·99
7' 7'49 17° 14·42 27' 26·51
So 8'02 - ISO 15'3~ 28' 28·10
9' S'57 19' 16'3 29' 29·78
30.0 31'S5
When the determinations are made rapidly and com-

parative values only are required (as in a tecp.nical analysis),
the correction for temperature and pressure is often omitted,
provided no appreciable-variations have occurred during the
analysis. It may be mentioned 'that a.variation of 7 mm. in
pressure or of aoe in the temperature introduces an error of
the order of I per cent in the percentage analysis of the gas.
Automatic correction to N.T.P. Various instruments have
been devised with the object of eliminating the above calcu-
lations of a volume of gas to N.T.P. One of the best known is
.Lunge's gas volumeter: this is IJarticularly useful where a
large number of volume measurements have to be made as,
BB
738 Quantitative Inorganic A 1Jalysis
for example, in the measurement of the volume of gas evolved
from a weighed quantity of soli51 or liquid. The device is
capable, however,. of general appli-
cation. The apparatus is shown in
- Fig. 6-7. A is the graduated tube
in which the gas is to be measured,
B i:) the" control" or " compensa-
tion " tube which has a spheri~al
or cylindrical bulb at the upper
end; the volume to the first mark
of the latter is 100 ml. , the re-
maining narrow portion of the tube
being calibrated up to 130 to 140
ml. in 0·1 ml. The" control" tube
is set once for all at the beginning of
B c the work by observing the tempera-
A
ture and pressure, and calculating
the volume which 100 ml. of dry (or
moi$t) air at N.T.P. would occupy
under these conditions. * The" con-
trol" tube is then filled to the
lTolume so calculated. if now the
levelling tube C is adjusted so that
the mercury in B is levelled on
this 100 ml. mark, and the meas~
uring tube A is adjusted so that
FIG. 6-7.
the mercury levels in A and B
are the same, the gas in A is under the same conditions
of temperature and pressure as that in B. The reading on B,
however, is lplown to represent the true volume of the gas in B
corrected to N.T.P., hence the volume read on the graduated
tube A will correctly represent Dhe volume at N.T.P. B is
kept dry or wet according as to wh~ther the gas in A is dry or
wet: the tubes' are kept dry or wet by means of a drop of
concentrated sulphuric acid or of water respectively. The
instrument should be used in a rOOql at uniform temperature,
• The necessary volumes V are given by the expressions:
100 (273 + t) 760
for dry gases V = ;
273 X P
. 100 (273 + t)'760
for mo~st gases V = ,
. 273 X (P - P)
where P is the atmospheric pressure expressed in mm. of mercury, t is the
temperature in degrees Centigrade, and p is the vapour pressure of water in
mm. of,.mercury at tOe.
Gas Analysis 739
and care must be taken to avoid heating one limb unequally
either by contact with the hands or by unequal exposure to
heat or draught. It may be mentioned that a difference of
9nly 1°C would mean a variation of 0·3 per cent in the gas
volume.
There are large numbers of automatic compensators, .but
for these the reader 'is referred to treatises ·on gas analysis.
VI, 5. Calibration of gas measuring vessels.-For work of
only moderate precision the calibration of gas burettes, etc., is
usually unnecessary; since instruments can be obtained with
fairly' accurate graduations. If, however, high accuracy is
required, or the graduation is su:;;pect, or the apparatus has
been made or repaired in the laboratory, calibration must be
carried out. Either the direct weighing m~thod or that
employing a " burette calibrator" (Section II, 22) may be
used. The liquid employed in calibration should be ,the same
liquid as is subsequently to be used -in the hqrette, I:!tc. In
the majority of cases mercury is employed. A cathetometer
is to be preferred fer viewing the meniscus. The temperature
of the mercury is, of course, noted, and the volume is calcu-
lated from the weight with the aid of Table XXV.
Meniscus correction. Gas bur-
ettes are usually calibrated inverted
(see Fig, 6-8). If we consider the
calibration of the burette in Fig. •
6-8 (a) with mercury (which forms
a convex meniscus), the volume
desired is up to the plane CD; it
is evident that the volume of mer-
cury weighed did not include the
space AEC and BED. Moreover
since the instrument is to be used
in the reverse position, the error is
really twice as great as is clear by
iI1spection of Fig. 6-8 (b). This is
taIled the double meniscus correction,
and will obviously depend upon
the internal diameter of the tube:
the correction will be positive. If
water is used in the burette for
analysis and calibration purposes,
the meniscus is concave upwards, (a) (b)
and the correction instead of being FIG. 6-8,
positive, as in the case of mercury, will be negative. Table
XXVII, due to Gockel (1903), gives the values for the
double meniscus correction with mercury and water res-
pectively for tubes of various diameters.
TABLE XXVII. DOUBLB MENISCUS CORRECTION FOR WATBR AND MERCURY.
Diameter
of tube
Double meniscus
correction, mi.
Diameter
of tube
I Double meniscus
correction, mI.
mm. Mercury Water mm.
I Mercury Water
1 0·0026 0·0021 14 O-US 0·291
2 0'0052 0·0043 15 0·140 0·343
3 0'0090 0·0093 16 0·162 0'395
4 0·0128 0'0144 17 0·174 0·443
5 0·0200 0·026 18 0·185 0·492
6 0·027 0·039 19 0·181 0·520
7 0·035 0·062 20 0·177 0'548
8 0·0~3 0·084 21 0·180 0·605
9 0·056 0·114 22 0·182 0·663
10 0·068 0·144, 23 0·179 0·708
11 0·080 0·190 24 0·176 0·754
12 0'091 0·236 25 0·173 0·800
13 0·104 0·263
VI, 6. Apparatus employed in gas analysis.-Numerous

forms of apparatus haye been designed for gas analysis, and
all that can be attempted ~ this Section
is to give a brief account of the construc-
tion of only a few typical forms.of apparatus.
A. Hempel's apparatus. This apparatus
was formerly extensively employed for
analise~ of an accuracy of about 0·5 per
cent. It consists of two parts, the burette,
which includes a measuring tube and a
levelling tube, and the pipette. The ~s
is usually measured over water or some
aqueous solution. '
The Hempel-Winckler gas-burette consists
of two glass tubes mounted on feet, one
tube being graduated to 100 ml. in 0·2 ml.
(the burette proper) and the other plain (the
pressure tube). They are connected at the
FIG. 6-9. feet by a long rubber tube as in Fig. 6-9.
Gas Analysis 741
Here the burette is shown' with a 2-way tap at the upper
end and a 3-way tap at the lower end (this is the Winckler
modification of the original Hempel bureUe, and must be
employed for gases which are
soluble in water). The original
Hempel burette contained no c
glass taps, but was closed by
a pincl::J.-cock on a rubber
tube at the upper end (this
is shown in Fig. 6-17 in
connexion with the method
0{ manipulation) ; the upper
end consists of a capillary
tube of I mm. internal and
6 mm. external diameter.
The glass tap burette is,'
however, more convenient
in practice and more
accurate.
A simple Hempel gas pipette

is shown in Fig. 6·10. This
consists of two large bulbs A
and B joined by the tube D ; FIG. 6-10.
C is a capillary V-tube of 1
mm. internal diameter and
about 6 mm. external diameter. Bulb A holds about 150 ml. and
bulb B about 100 ml., so that when 100 ml. of gas is brought
into A, sU.fficient space for the absorbing liquid will remain.
The pipette is securely fastened to a wooden or iron stand;
an iron stand with a four-sided base is generally preferable
to wood because (i) its greater weight renders the apparatus
more stable and {ii) it canciot warp. This pipette is employed
for submitting the gas originally present in the burette to the
ac~ion of a special liquid absorbent. The bulb A is filled
with the absorbent liquid by pouring the Jatter into Band
then. sucking the air out of the apparatus at the capillary C ;
the liquid should rea,.ch the siphon bend of the capillary tube
and bulb B should be nearly empty. It is convenient to keep
a number of these pipettes filled with various absorbents
(see Section VI, 7), well corked (or closed by rubber" police-
men ") and labelled with the date on which fresh reagent was
intr01uced. On each occasion that the pipette is used, the
74.2 Quantitative Inorganic Analysis
fact is noted -on the label; in this way it is possible to know
approximately when it is
C necessary to recharge the
apparatus.
A pipette suitable for solid
(and liquid) reagents. such as
phosphorus in water, is shown
in Fig. 6-11. The solid re-
agent is introduced through
D, which is then closed with
a rubber stopper and securely
wired into place. This pip-
ette is also useful for liquid
reagents (e.g., potassium
hydroxide for carbon dioxide
determinations) for it may
be filled with rolls of wire
gauze or with fine glass rods
before filling'the pipette with
the reagent; the gas is thus
exposed t9 a larger surface
of/the reagent and absorption
FIG. 6-11. is more rapid.
For solutions which

undergo change upon ex-
posure to the air, such as
an alkaline solution of
pyrogallol or an ammon-
iacal solution of cuprous
chloride, the double "absorp-
tion pipette (Fig. 6-12) is
employed. Here the re-
agent in B is in contact
with an atmosphere free
frum oxygen, the indiffer-
ent gas being confined
over water in the bulbs
C and D. This pipette is
filled as follows. Tube E
is first closed with a
rubber stopper, and the
apparatus is filled with FIG. 6-12.
Gas Analysis 743
hydrogen or nitrogen by passing the gas through it. The
reagent is introduced by pouring it through the tube E, allo
FIG. 6-13.
air-free water in through G until the bulb C is almost fuU-

in order to form a water seal. The tube E is then closed by
a rubber stopper, and wired on.
Fig. 6-13 is a double absorption
pipette but adapted for use with
a solid (or liquid) absorbent.
Fig. 6-14 shows the simple
absorption pipette used for fuming
sulphurj.o acid. This has an
upper small bulb, about 5 cm.
in diameter, filled with glass
beads or broken glass in order
to increase the absorbing sur-'
face. The ends of this pipette
are closed- when. not in use by
small glass caps, which can be
made air-tight with rubber
bands; a considerable im-
provement is effected by the
use of groun~-glass caps. FJG. 6-14.
Fig. 6-15 depicts a Hempel explosion pipette; this contains
mercury instead 01 water. A slow combustion pipette is
illustrated in Fig. 6-16. The latter consists of a Hempel
FIG. 1i-15.
FIG. 6-16.
pipette for solid reagents with iron (or nichrome) electrodes

inserted through the tubulature; the electrodes te:t:minate in
a platinum spiral P, which is heated by a current of 2 to 3
amps. at 6 volts.
Gas Analysis 745
Manipulation of the Hempel apparatus. The complete
apparatus is shown in Fig. 6-17. The ex;perimental details
for a simple Hempel burette
(with pinch-cock) will be
given; the student should E
have no difficulty in modify-
ing these to the use of the
Hem pel-Winckler burette.
I t is recommended that the
pinch-clamp at E be of the
Day type, * as this is far
superior to the simple
pinch:-cock usually employed : B
with burettes.. If the bur- A
ette has been in use, clean
it thoroughly in the usual
manner (Section n, 19), and
rinse it welt with distilled
water. Make sure that the o
thick rubber tubing (" pres-
sure tubing") joint at the
upper end of the burette is
securely wired.t Open the FIG. 6-17.
pinch-clamp at E, and slip it
down over the capillary tube. Pour water that has been
saturated with the gas mixture to be analysed into the burette
until A and B are rather more than half full. Drive out all
the air from the rubber tubing connecting A and B by raising
and lower.ing the tubes alternately, keeping the rubber tubing
ta-ut during the process. Raise A until the water begins to
flow out of the top of B, then close the rubber tube of the
bur'ette with the pinch-clamp, which is placed close to the
end of the capillary. Compress the rubber tubing at C
between the thumb and fihgers of the left hand, and pour out
- the excess of water that is in A.
* Supplied by the Central Scientific Co., Chicago. Ill.. U.S.A., and also by
E. H. Sargent and Co., 155-165 East Superior Street, Chicago. Ill., U.S.A.
t When joining together glass tubes by rubber tubing, the ends of the tubes
should be rounded off in the flame, and then brought close together within
the piece of rubber tubing. ,The rubber tubing is secured in place by means
of wire ligatures, composed of copper wire about 1 to 2 mm. in diameter.
Each ligature should consist of only one turn bf the wire around the tube, the
ends being drawn and tightly twisted together by means of a pair of pliers.
Long rubber connexions should be avoided, not only because the rubber
tubin~ is somewhat porous, but also because air in the tube tends to adhere
tenaclously to the walls.
BB·
The next step is to measure out exactly 100 ml. of the gas
sample. Insert into the rubber tube at E a capillary tube
connected with the supply of the gas to be analysed, after
first displacing the air in this connecting capillary by the gas.
Hold tlie upper part of A in the left hand, lower,it belOw B,
and open the pinch-clamp E with the right hand. Introduce
somewhat more than. 100 ml. of gas into the burette. Close
the pinch-cock E, level A and B, and allow the water in the
burette to drain for 2 minutes. Disconnect the capillary tube
from the source of gas, and adjust the volume of the gas by
suitably :tnanipulating the pinch-clamp until it is exactly 100
ml. when the levels of the liquids in the two tubes are the
same.
The next stage is the absorption of one of the gaseous com-
ponents of the mixture. Connect the burette and absorption
pipette exactly as in Fig. 6-17: the burette and level tube
should be on a flat table. Insert into the rubber tubing G
the bent capillary tube F. This connecting capillary has the
same dimensions as the capillary tubes of the burette and
pipette, viz., 6 mm. external, and 1 mm. internal diameter;
the horizontal portion is about 6 em. lon~ and the legs each
about 2'0 cm. long. Slip a long piece of rubber tubing over
the tube H of the pipette. Grasp F between the thumb- and
fingers of the right hand, squeeze the rubber tube E on the
capillary tube of the burette between the thumb and fingers
of the left hand, blow gently through H until· the liquid in the
pipette is driven over to a mark on the farther end of the
horizontal portion of the capillary tube F, and then insert the
end of F into the rubber tube on the burette. If the con-
nexion is properly made, there will be p.ractically no move-
ment of the reagent in the capillary. Even if the air enclosed
in the connexion between the burette and pipette should
.occupy about I cm. length of the capillary, the error may be
neglected since capillary tubing of about 1 mm. bore occupies I
,only about 0·01 ml. per cm. length. If; however, more than
1 cm. length of air remains in F, F should be slipped out of E
and the operation repeated.'" The apparatus is now ready
for the absorption of a gas.
Open the pinch-clamp E, and slowly raise the levelling tube
.* Certain reagents must never be allowed to corne into contact with the rub-
ber connexlon at G; in such cases. tbe liquid from the pipette is forced up-
wards until it stands just below this rubber connexion and allowance is made
for the air remaining in the capillary tubes. The correction is readily made
since 10_cm. of the 1 mm. bore capillary corresponds to 0·1 ml. of air.
Gas Analysis 747
A, thus driving the gas over into the pipette. Allow the
water to flow into the capillary tube F until it reaches the
point to which the reagent has been driven over, and then
close the pinch-clamp. A thin layer of the reagent adheres
to the walls of the capillary, and this is sufficient to remove
most of the absorbable constituent of the gas mixture remain-
ing in the capillary. With some gases (e.g., carbon dioxide)
absorption occurs rapidly in the pipette, but with others (e.g.,
oxygen and carbon monoxide) it is necessary to shake the
pipette after the gas has been transferred into it. In the latter
cas~, place another pinch-clamp on G, hold the burette in the
left hand and gently rock the pipette stand backward and
forward on the front edge of the base. After the c<?nstituent
that is to be removed. is completely absorbed, the gas is·
transferred back into the burette. This is don~ by holding
the levelling tube in. the left hand and in such a position that
the confining liquid in it stands at a slig·htly lower level than
in the burette, open pinch-clamp E (and, if necessary, that at
G) and lower the levelling ,tube slowly until the gas is drawn
back into the burette and the liquid in the pipette reaches a
point on F at which it originally, stood; close the pinch-clamp
E, level approximately, allow the water to drain for 2 minutes,
level accurately, and take the reading of the volume of the
gas in the burette. Absorption may not be complete in one
operation. It is therefore advisable to transfer the gas back
to the absorption pipette and to repeat the process until no
further diminution in volume occurs.
By the use of a series of Hempel pipettes the gas is sub-
jected to the action of various absorl!ents, one after another,
until the complete composition is ascertained. As already
pointed out, the connexions must in all cases be made with
stout rubber tubing (" pressure" "tubing) which are secured
with copper wire ligatures.
For gases which are appreciably soluble in water, the
levelling tube containing water cannot, of course, be used.
The clean burette, which.. must be of the Hempel-Winckler
type, is first, thoroughly dried by washing it successively with
alcohol and ether (alternatively by acetone alone) and then
drawing air through it. The dry burette is filled with the
gas either by the pressure of the gas or by using an aspirator
bottle attached to the 3-way tap to draw out the air and
replace it with the gas; when the burette is full, the taps are
turned off, care being taken that the gas is at atmospheric
pressure. Connexion may then be made with the levelling
tube A. which is then fin~d with water, th~ lower tap opened
so that the water flows into the burette and absorbs the
soluble gases present. As the burette holds exactly 100 ml.
between the 3-way tap and the upper tap, the percentage of
soluble ga.s can be read off directly.
The manipulation of the Hempel apparatus may be con-
siderably simplified by utilising a special Hempel burette as
shown in Fig. 6-18, and extending the capillary tube of the
pipette so that it has
a horizontal arm. The
burette has a special
3~waytapwithcapiltary
tube arms; only one
.rubber connexion is r~
quired for attachment
to the pipette.
A-ttention is directed
to the fact that the
Hempel apparatus has
been modified for the
FIG. 6-18. exact analysis of gas- ~
eoos mixtures chiefly
by (a) the use of a special burette with a device -for
automatically compensating for yariations in temperature
and barometric pressure, and (b) by the use of mercury as the
confining liquid; the' latter necessitates that the bum B
(see Fig. 6-10) is attached by means of " pressure" tubing
to the bulb in which the actual absorption takes place. *
It is doubtful whether ~mch apparatus now finds a wide use
because of the introduction of other more convenient appara-
tus for precision gas analysis (see below).
B. Bunte's a~paratus. 'The apparatus is shown -in Fig.
6-19 (a). The absorbent is introduced directly into the
apparatus, and absorption takes place in the measuring vessel
itself. Analyses are only' of a second order accuracy, but the
apparatus is of great value for the preliminary. training of
large classes of students in gas manipulation.
The Bunte burette. in an improved for.m, has a capacity
of 100 mI. from the zero graduation to the upper 3-way tap,
and is also graduated downwards from zero to 10 to 15 m!.
for measuring the absorbing solution, if 'desired. In one form
* For further details, see L. M. Dennis and M. L. Nichols, Gas Analysis,
1929 (The Macmillan Co.).; G. W. Himus, Fuel Testing, 1932 (Leonard Hill).
Gas Analysis 749
the tube is of uniform bore and is graduated in 0·2 mI.
divisions;
, in another form (not shown in the Figure) there is
(a) (b)
FIG. 6-19.
an upper bulb (capacity 50 ml.) graduated in 1 ml. divisions,

and the lower half of the apparatus (50 ml. + 10-15 ml.) is
the ordinary cylindrical 'tube graduated in 0·2 ml. C is a
small cup-shaped vessel, which is somewhat over 25 ml. in
capacity, and is provided with a 25 nil. (and sometimes also a
20 ml.) graduation mark. For more accurate results, the
burette is surrounded by a wa.ter jacket, but this is ordinarily
omitted. The capillary below B may be connected by means
of rubber tubing to a levelling reservoir or bottle containing
water; 'this is adjustable to any height and is supported on a
small circular platfo"i-m fitted to a tall burette stand. Fig.
6-19 (b) is a gas wash bottle, fitted with short lengths of.
" pressure" tubing at F and G, which can be closed by means
of screw clips or pinch-clamps.
Manipulation of the apparatus. Fill the levelling bottle,
preferably saturated with the gas to be analysed, and connect
it to the burette below B. Open taps B and A, and allow the
water to run up to the upper mark in the cup C. Connect
the stop-cock A through D to the supply of the gas, open B,
lower the levelling bottle, turn tap 1\ to the proper position
and draw 101 to 103 ml. of the gas into the burette, and close
tap A. Raise the levelling bott~e, and compress the gas until
the water exactly reaches the zero mark on the burette;
close tap B, open tap A cautiously so that the excess of gas
bubbles through the water in the cup. 'The gas in the burette
is now under atmospheric pressure plus the pressure of the
water in the cup C; these conditions must be reproduced in
all subsequent measurements. Place a screw clip or pinch~
clamp on the rubber tubing connecting the burette to the
levelling bottle, and pull off the end of the tUbing attaclied to
the burette.
Almost fill the wash bottle, Fig. 6-19 (b), with water.
Blow up the wat~r into the rubber tube F, and close it with a
screw clip. Attach the upper end of F to the lower end of the
Bunte burette. Open tap B. and connect the wash bottle to a
filter pump (at G): open the screw clip at G~ and suck t)le-
water in the burette until it is just above the tap B, which is
then closed. The gas is now under less than ..atmospheric
pressure, and when the tip of the burette is dipped under the
absorbent in a shallow dish and the tap B opened, the absorbent
will rise in the burette. Disconnect the wash bottle, place
the end of the burette below the' surface of a suitable absorbent
contained in a shallow dish, and allow a suitable volume to
enter; then close tap B. Hold the burette above the upper
tap and below the lower tap (to prevent heatIng of th.e gas),
and shake well for a few minutes. Dip the tip of the burette
under the absorbent, open tap B, and allow a little more -of
the abso~bing solution to enter. Repeat the process untif
. no more of the absorbent is sucked up into the burette.
Empty most of the water from the wash bottle, and
arrange so that 'water is contained only in the central
tube and the rubber tubing F. Attach the water bottle to
the lower end of the burette and connect G to a filter'pump ;
Gas Analysis 751
suck out the absorbent from the burette until the absorbent
just reaches tap B, and then close the tap. Dip the end of the
burette into a dish containing water, which then rises in the
tube. Close tap B, open tap A so that water flows into the
burette from the cup C until the original pressure is produced.
Add more water until it stands at the upper mark in the cup,
and open tap A cautiously to reproduce the original pressure.
Read the volume of the residual gas: the difference in vo1ume
gives the concentration directly in percentage by volume of
the constituent absorbed.
The Bunte burette may be employed inter alia for the
determination of carbon dioxide (by aqueous potassium
hydroxide), heavy hydrocarbons (by bromine water), oxygen
(by an alkaline solution if pyrogallol or, better, by alkaline
sod.ium hyposulphite solution), and carbon monoxide (by
ammoniacal cuprous chloride solution) .
. C. Orsat's apparatus. This apparatus, introduced by
Orsat in 1874 and since modified by numerous workers, is
widely used where an accuracy of about 0·5 per 'qmt suffices,
e.g., for furnace gas analysis. The method of analysis is
essentially similar to that with the Hempel apparatus, but
the burette, pipettes and accessol'ies are so designed that
the wbole apparatus, including three or four absorption
pipettes, fit into a portl1ble box.
The Orsat-Lunge app,aratus is sp_own in Fig. 6-20. It
consists of a water-jacketed burette of 100 ml. capacity,
attached by means of a capillary tube manifold to four
pipettes, three of which contain absorbents for carbon diox;ide,
oxygen and carbon monoxide respectively and which are
usqally filled with glass tubes or rods to present a greater
surface of the reagent to the gas; the fourth pipette contains
water only and is used in conjunction with a tube containing
purified asbestos impregnated with palladium and heated by
means of a small spirit lamp. The last pipette and palladium-
asbestos tube is omitted in some forms of the apparatus.
The burette and reservoir to which it is attached contain
acidulated water coloured with a suitable indicator, so that an
indication is" given at once of any accidental fouling with
reagents. The gas mixture is drawn into the apparatus
through the filter shown on the outside of the case (the filter
may be dispensed with when working with gases free from
dust) and through the 3-way stop-cock. If the gas to be
sampled is at less than atmospheric pressl.lre, the sampling
tubes and filter are filled with gas by operation of a rubber
aspirator (not shown in the figure) before the stop-cock is
FIG. 6-20.
..
co.nnected with the measuring burette. The first sample of
gas drawn into the burette is discarded, it being employed to
expel the air from the capillary connecting tubes. The
"reagent in each pipette is adjusted to a standard line just
Gas Analysis 753
below the rubber connection, and under no circumstances •
must any of the reagent be drawn above the tap.
The method of carrying out the absorptions for carbon
dioxid~, oxygen and carbon mQnoxide is obvious (it is similar'
to that employed with the Hempel apparatus), and need not
be further described. For the determination of hydrogen,
which must be carried out after the above gases have been
-removed from the gaseous mixture, air is allowed to enter the
burette until the total volume is approximately 100 ml. and
the exact volume is read. The small capillary containing
palladinised asbestos (situated at the left of the bulb con-
taining water) is gently heatM by means of a small spirit
lamp, and the gas is passed very slowly from the burette into
the pipette: the hydrogen will combine with the oxygen at
the surface of the catalyst. The gas is returned to the burette
and the process repeated, after which the volume of the resi-
du~l gas is measured. Two-thirds of the contraction repre-
sents the volume of hydrogen. If the temperature is not
allowed to rise above 200°, the volume of any methane and its
homologues present which is burnt may be neglected in
technical analyses. Various improvements for the deter-
mination of hy,drogen and methane, either alone or in admix-
hire, have been introduced. The most important of these are
(i) a Pyrex U-tube containing cupric oxide' and which can be
heated in a small electric furnace at 270° to 300°C (fractional
combustion "pipette "). and (ii) a pipette provided with a
platinum wire spiral which can be heated electrically and its
temperature controlled by a rheostat (slow combustion pip-
ette). 'The fractional combustion pipette is employed to
oxidise both carbon monoxide and hydrogen: at 270° to
300°C the oxidation may be effected in the presence of satur-
ated hydrocarbons, such as methane and ethane (compare
Section VI, 8). The diminution in volume gives the volume of
hydrogen directly, whilst the combustion of carbon monoxide
produces an equal volume of carbon dioxide (2CO + O2 =
2C02), which can be determined by absorption; in the latter
case mercury must be employed as the confining liquid. As
• a general rule in the analysis of mixtures containing methane
and ethane, it is best to determine the carbon monoxide by
absorption and t~e hydrogen by fractiqnal combustion over
cupric oxide at about 290°C. The gas is passed through the
copper oxide U-tube maintained at 290 0 to 300°C at a rate
of 10 ml. per minute and back into the measuring vessel, and
the process is repeated until there is no further contraction,
754 Quan#tative Inorganic Analysis
, 'The slow combustion pipette with, pr~ferably, mercury as the
confining liquid is employed for the determination of methane
arid/or ethane. A commercial form of portable apparatus is
available which incorporates absorption pipettes of improved
design, a fractional combustion" pipette," and a slow com-
bustion pipette*; this apparatus can therefore be employed
for the determination of carbon dioxi~e, unsaturated hydro-
carbons, ~arbon monoxide, hydrogen, methane, ethane, and
nitrogen (by difference).
A serious drawback of the Orsat-Lunge and similar appara-
tus is that the absorption of oxygen and carbon monoxide is
nut complete unless the absorbent and gas are shaken
together or unless the gas is brought into prolonged and inti-
mate contact with the absorbent. Shaking is, of course, not
very practicable with this apparatus; however, the use of
pipettes packed with glass tubes considerably improves the
absorption, but is nevertheless not quite
satisfactory. Various improvements in the
design of pipettes have been introduced, for
example, by Dennis (spiral type)t and by
otherst, but perhaps the best of these is
due to M. Shepherd (1930). The latter is
shown in Fig. 6-21** and is of the bubbling
type the gas passes through a capillary
tube, at the lower end of which is a plat-
inum gas distributing disc (containing about
200 individual minute perforations) is sealed.
The gas is thoroughly atomised, intimate
contact is thus ensured, and absorption is
complete.
The Orsat apparatus has been so modified
and improved by M. Shepherd (1931) that
it may be employed as a precision instru-
ment for the analysis of gaseous mixtures.
Mercury is employed as the confining liquid.
The Shepherd apparatuS' in Pyrex glass is
FIG. 6-21. available commercially§; for further details
• This is the " Fisher" gas analyser, technical universal model; made by
the Fisher Scientific Co., Pittsburgh, Pa., U.S.A.
t L. M. Dennis and M. L. Nichols, Gas Analysis, 1929, p. 76.
t See the apparatus catalogues of the Fisher Scientific Co., E. H. Sargent
and Company, Baird and Tatlock (London) Ltd., A. Gallenkamp and Co.
Ltd., etc .
•• Manufactured by the Fisher Scientific Co., Pennysylvania, Pa., U.S.A.
§ Supplied by E. H. Sargent and Co., Chicago, Ill., U.S.A.
Gas Analysis 755
the reader is referred to the booklet supplied by the manu-
facturers and -also to the original memoir in the Journal oj
Research oj the Bureau oj Standards (1'931, ~, 121-127).
D. Ambler's portable apparatus. The apparatus, designed
by H. R. Ambler (1931), is portable (the size complete in case
is 15 X 12 X 4·5 inches and the weight is 8·5 lbs.), inexpen-
sive, and can be used either for p'recision work or for routine
determinations. * Mercury is employed as ·the confining
liquid. The instrument is eminently suitable for the training
of students in gas analysis.
The chief details of the apparatus are shown in Fig. 6-22,
whilst Fig. 6-23 is a line diagram of the actual apparatus.
Essentially the apparatus consists of (i) a 10 ml. burette A,
surrounded by a water. jacket, in which the gas is measured,
(ii) a vessel B in which the gases are absorbed or combusted,
and (iii) a graduated mercury manometer M. A and B
are connected to mercury reservoirs Rl and Rz respectively;
Tl and Tz are 3-way and 2-way taps respectively; reference
marks are etched at F and K; B contains a thin platinum
coil, made from 1·5 cm. of platinqm wire of 1·25 mm. diameter,
and current is obtained from a 4·5 volt dry battery which is
fixed on the base of the instrument and is controlled by a
3-ohm rheostat of the wireless component type; the reservoir
Rl is provided with two rests (shown in Fig. 6-23~ at levels
near the top and bottom of the manometer scale.
The general mode of use of the apparatus is a~ follows:
1. Introduce the gas sample (up to -10 ml.) into the vessel B
by means of the capilfary tube H, the tap Tl and the reservoir
·R2•
2. Reverse the tap T 1, and transfer the gas to the measuring
vessel A. Run the mercury to the mark K, and thu§ seal the
gas in the burette.
3. Introduce a quantity of the appropriate absorbent into
B from:a by suitably manipulating Tl and R 2•
4. Transfer the gas to B with the aid of Rl and T 2 • Ab-
sorption will take place.
5. Transfer the residual gas back into the measuring burette
A by raising R2 and suitably turning T 1 • As soon as the
absorbent liquid has reached the left-hand side of the tap T l'
reverse the latter, and drive out the reagent through H.
• The apparatus, and also a "work~ model" .(with manometer omitted),
is supplied by A. Gallenkamp and Co. Ltd., Sun Street, London, E.C.2,
England.
Then reverse the tap, and adjust the mercury to the mark K
as before. •
The method of measutement of the gas varies according as
to whether precision results (within 0·1 per cent) or moder-
FIG. 6-22.
ately accurate (" technical," i.e., within 0·5 per cent) are
required. The measuring vessel consists of a burette of 10
ml. capacity, calibrated in 0·1 ml. Below the 10 ml. mark
the burette narrows to a bore'tlf 3 mm., and a mark F is
etched round the lower part. Either varying volumes or
varying pressures ,may be measured; the temperature is
assumed constant, but may be corrected for if variations
occur. The constant volume method is the more accurate.
Three procedures will be described.
Gas Analysis 757
(a) Varying volume and constant pressure. This is the
qUlckest procedure and yields results of "technical" accur-
acy. Level the mercury,in the burette with the mercury in
R1 , which is held in the hand, and read the volume.
FIG. 6-23. •
.' (b) Varying volume and approximately constant pressure

(corrected). This modification of the constant pressure
method is the most satisfactory for precision work. Bring
the gas in the burette to approximately atmospheric pressure
(say, to within 2 mm.), close the tap T 2 , ~nd read both the
burette and manometer. The pressure in the burette is then
equal to the atmospheric pressure, plus the difference between
the manometer reading and the reading of the maRometer
when the burette is filled to the same mark at atmospheric
pressure. Let us call this difference AP. I
The manometer readings corresponding with the different

burette readings at atmospheric pressure are a constant of the
apparatus, and are determined once and for all by measuting
a few points with the tap T 1 removed. These are plotted in
the form of a graph, whicli is a straight line.
The correction to the burette reading is applied as follows.
Let P be the atmospheric pressure, p the vapour pressure of
water, V the burette reading (i.e., the volume of moist gas at
the experimental pressure), and Va what the burette reading
would have been if the mercury has been levelled to exactly
atmospheric pressure. Then:
V - V P - P + L1P,
,,- X P_p
or V - V= V L1P.
" X P _po
If L1P is not greater than 1 cm., then any deviation of P - P
from the average approximate figure (76 - 1) = 75 cm. will
not affect the correction sufficiently to introduce an error of
more than 0·1 per cent on the total gas: In practice, L1 P
does not exceed 2 mm; the magnitude of the correction, for
1 mm. of L1P, is :
V X (0·1/75) = V X 0'0013, i.e., 0·13 per cent of V.
(c) Varying pressure and constant volume. Many gas
samples which occur in practice do not involve corrections
greater than about 30 per cent of the total gas; for this
purpose a manometer of 30 cm. length will suffice. In such
cases the maximum accuracy is obtaine.d by the constant
volume method. A knowledge of the v.olume is not required,
and the result will be independent of any errors in. the
graduation of the burette. The mark F serves as a levelling
mark for this purpose,~nd .the bulb below the water jacket
enables a sufficient volume of gas to be taken in the measuring
burette at atmospheric pressur~ to be compressed to ca.
10·5 ml. (volume to mark F) at about 100 cm. pressure. It
can be shown that the volume of the gas (at normal pressure)
contained in the b~rette is proportional to :
P + A - K - P /76,
where A is the manometer reading and K is the manometer
reading.with the burette at atmospheric pressure (all pressures.
are expressed in em. of mercury).
Any other graduation may, of course, be empToyed, as in
the case of samples which ,are not large enough to fill the
burette. The precision then is somewhat less, since any error
in levelling in the wider tube causes a greater error in volume.
Mention must be made of the fact that the vessel B may be
fitted either with a spark gap or with. a platinum spiral for the
burning of hydrogen ana other combustible gases; the plati-
num sEiral is of greater general utility. Gas mixtures con-
Gas Analysis 759
taining more than 10 per cent of hydrogen or 3 per cent of
hydrocarbons should be diluted with air before combustion.
The apparatus is applicable to all gases commonly dealt
with. It can be used equally well for precise general analysis
and for rapid technical analysis, and the volume of the sample-
required is small (normally 10 mI.). The use of mercury as
the confining liquid increases the speed of transference of the
gas, and determinations are correspondingly rapid.
E. Bone and Wheeler apparatus. This, in its simple or one
of its modified forms, is probably the II).ost widely used instru-
ment for exact analysis
in Great Britain, and
is applicable to most-
classes of gaseous mix-
tures. The usual form
of Bone-Wheeler appar-
atus is shown in Fig.
6-24. The gas mixture
is measured at constant
volume in limb A of the
water-jacketed U-tube,
the pressure b e i n g
measured by the height
of the mercury column
in .the limb B, which is
graduated upwards in.
millimetres. The limb
A has a series of grad-
uations coinciding in
level with each integral
100 mm. mark ori B.
The heights pf A and
B are such, that when
a measl,lrement is be-
ing made, the closed
space above the mer-
cury in B is a Torri-
cellian vacuum except
for water vapour ; both FIG. 6-24.
limbs are kept satur-
ated with water by means of a small quantity of dilute
sulphuric acid {2 to 5 per cent). The pressure of the gas
in A is equal to the diff~rence in the level of the mercury
760 Quantitative I norgant'c Analysis
in Band A. Variations in the atmospheric pressure during
the· analysis do not affect the results, but variations in tem-
FIG. 6-25.
Gas Analysis 761
perature do affect the readings and corrections must be
calculated (see Section VI, 4). Mercury is raised or lowered
by means of the reservQir D, which is connected ,top. windlass
and ratchet; fine adjustments inJevelling on to the marks is
obtained by means of a special tap C, which separates the
limbs A and B from D.
Abs9rption is carried out in the vessel (or pipette) F, which
stands in a mercury trough; the capillary 3-way tap at the
top connects to the measuring vessel and to a bottle or other
container connected to a filter pump or other exhausting
apparatus, by means of which the reagent, introduced into F
by means of a special bent pipette, may be removed after
absorption has taken place. Combustible gases are deter-
mined by explosion with air i.n the vessel K, which is provided
with a spark gap.*
Various modifications nave been introduced, notably by
J. G. King (1922) who incorporated a copper oxide tube for
combustions,t and by Bone and Newitt (1929). The Bone
and Hewitt apparatus is shown iri Fig. 6-25:1:; the chief
distillctive feature is the provision of a 3-way water tap of
special design, by means of which the mercury levels are
controlled from the water supply with ease and delicacy.
The levels are read with a cathetometer. Oth~rwise, the
procedure is similar to that with the Bone and Wheeler
apparatus. **
A precision apparatus which permits of the analysis. of 10
ml. of gas with an accuracy of ± 0·02 per cent has been des-
cribed by B. Lambert and D. J. Borgars (Phil. Trans., 1939,
237, 54;1-565).
F. Other precision forms of apparatus.-Most of these are.

of the" absorption pipette" type and employ mercury as the
confining liquid. They are.based upo"n the design originally
introduced by Burrell and Oberfell (1916) and are 6ften spoken
of M U.S. -Bureau of Mines apparatus. The SheRherd appara~
tus (Section VI, 6C) belongs to this class. Commercial forms
.. For further details. see G. Lunge-H. R. Ambler. Technical Gas Analysis.

1934 (Gurney and Jack!,\on); A'. McCulloch, Gas Analysis, 1938 (Witherby) ;
F. Sutton·A. D. Mitchell, Volumetric Analysis, 1935 (ChurChill).
t See the catalogues of Baird and Tatlock (London) Ltd., London- and ot
A. Gallenkamp and Co. Ltd., London, England.
t Fig. 6-20 is reproduced by courtesy of A. Gallenkamp and- Co. Ltd.
** For complete details of the manipulation of the apparatus. see G. W.
Himus, Fuel Testing, 1932, p. 185 (Leonard aill) ~
of these gas analysers are marKeted by several manufacturers. *
VI, 7. Absorbents for the various gases.-With all the types
of apparatus described in the previous Section, the methods
of determination of the constituent gases are essentially
similar. The main sources of error by absorption methods
are: (a) incomplete absorption .due to insufficient time of
contact, the use of an exhausted solution, or a chemical
equilibrium between the gas and the compound formed DY its
action on the absorbent; (b) physical solution of constituent
gases in reagents designed for the absorption, of other con-
sHtuents. The latter error is reduced to a minimum by Hie
use of as 'small quantities of reagents as possible, and by
saturating the reagents with the gases with which they come
into contact: this is particularly necessary with instruments
of the Hempel and Orsat type, where small volumes of
reagents are not practicable. •
The component~ of the gas mixture are usually determined
in the following order by the methods detailed below.
A. Carbon dioxide. The most common reagent is a 25 to
33 per cent aqueous solution (by weight) of A.R. pota~sium
hydroxide. For use in the Orsat apparatus a 40 to 50 per
cent solution is recommended in order to reduce the number of
times necessary to refill the apparatus. Sodium hydroxide
should not be empl.oyed owi~g to the tendency of the bicar-
bonate to crystallise out; the ,potassium salt is much more
soluble'. .
It will be appreciated that acid gases, such as hydrogen
sulphide, sulphur dioxide and hydrogen cyanide, are also
absorbed by the alkaline solution, and must be removed first.
. B. Unsaturated hydrocarbons. The most satisfactory
absorbent is fuming sulphu~ic acid, .con~aining 20 to 25 p~r
cent of free S03; the spectfk gravlty IS about 1·94. ThIs
absorbs all 'the olefines, acetylenes and aromatic hydrocar-
bons. The reagent must be employed at temperatures above
15°C, since Below this temperature pyrosulphuric acid may
crystallise out. Absorption is usually complete in 2 to 5 min-
utes; 'the rate of absorption is greatly accelerated by increas-
ing the surface of the reagent by ihe use of glass rods or
spheres (compare Fig. 6-14). The gas must be subsequently
·.For further details, see, for example: (i) Dennis and Nichols. Gas Analysis
(1929); (ii) the apparatus catalogues of E. H. Sargent and Co., Chicago,
U.S.A., and of Fisher Scientific Co., Pittsburgh, Pa., U.S.A,; (iii) the-booklet
of t)le·last-named firm entitled Fisher Gas Analysis Manual (revised edition).
Gas Analysis 763
treated with potassium hydroxide solution to absorb acid
vapours. Fuming sulphuric acid must not be kept for long
periods over mercury, although it is quite safe to allow it to
remain in confact with mercury for a few minutes, as in the
Bone and Wheeler absorption pipette. -
A solution of bromine in 10 per cent potassium bromide
solution is also employed to absorb unsaturated hydrocarbons.
Bromine water attacks mercury much more readily than
oleum, and is to be avoided, where possible, if mercury is
used as the confining liquid for the gas. It is said (Himus,
1932) that the concentration may be so adjusted (to a deep
yellow solution) that it absorbs effectively whilst having little
action upon mercury. The gas must be washed sub?equently
with alkali to remove bromine vapour.
In technical practice the heavy hydrocarbons are usually
absorbed together, but they may be absorbed separately by
suitable means.
(i) Acetylene. This gas is absorbed by ammoniacal cuprous
chloride solution with the formation of cuprous acetylide
CU 2C2 , but the reagent cannot be used in tIie presence of car-
bon monoxide and oxygen. With potassium mercuric iodide
reagent only acetylene is ab~orbed and carbon monoxide is
unaffected. The latter reagent (1 ml. absorbs 20 ml. of
acetylene) is prepared by dissolving 25 g. of mercuric iodide
and 30.g. of potas~ium iodide in 100 ml; of water, and mixing
it wi(h potassium hydroxide solution immediately. before use.
'(ii) Benzene. Benzene is readily absorbed by concentrated
SUlphuric acid, and. can thus be separated from ethylene
(which is not appr~ciably p.bsorbed), but not from the higher
olefines. B_enzene may be removed: by ammoniacal nickel
cyanide reagent; ethylene is not absorbed.
The ammoniacal nickel cyanide reagent is prepared as follows. Add a
solution of 25 g. of potassium cyanide in 40 ml. of watjlr to a solution of 50 g.
of crystallised nickel sulphate in 75 ml. of water. Introduce 125 ml. of con-
centrated ammonia solution into the mixture, and shake until the nickel
cyanide dissolves completely. Cool to O°C, and allow to stand at this.tempera-
ture for 20 minutes. Decant the clear liquid from the potassium sulphate
that has separated, and treat with a: solution of 18 g. of ctystallised citric acid
in 10 ml. of water. Allow the mixture to stand at O°C for 10 minutes, decant
the greenish-blue supematant liquid, and transfer to the absorption pipette
(e.g., a Hempel double absorption pipette, Fig. 6-13, with the large bulb filled
with broken glass). Add 2 drops of pure benzene to the reagent and shake
the pipette until the benzene has combined with the reagent (2 to 3 minutes) ;
this is shown by the appearance of fine, white, granular precipitate. The
additiqn of a little benzene is necessary because it has l:1een found that the
freshly prepared reagent does not rapidly remove benzene vapour until some
of the compound between benzene and the ammonium nickel cyanide has been
formed.
c. Oxygen. The three ;absorbents which are commonly
employed for the absorption of oxygen are a strongly alkaline
solution of pyrogallol, a solution of sodium, hyposulphite,
and phosphorus respectively.
Alkaline pyrogallol. For the Orsat and Hempel type of
apparatus a solution of [5 g. of pyrogallol in 100 m!. of 50
per cent aqueous potassium hydroxide (prepared from A.R.,
i.e., electrolytic, KOH) is satisfactory. For the Ambler and
Bone-Wheeler type of apparatus 40 per cent _potassium
hydroxide solution and 25 per cent aqueous pyrogallol solu-
tion are mixed in the ratio of 4 : 1 as required. At 20°C about
3 minutes' .shaking is required for complete absorption;
below 15°C absorption is inconveniently slow .. With gases
containing more than 25 p.er cent of oxygen, s_ome carbon
monoxide is always evolved, thus giving low results. A little
carbon monoxide is also stated to be evolved with certain
alkaline solutions of pyrogallol, but Haldane '(1933) has shown
that with the solution recommended above for the Orsat type
of apparatus the effect is negligible if the solution is kept for
3 days before use or, alternatively, is heated at 100°C fot an
hour.
Sodium. hyposulphite. The reaction which occurs is:
Na 2S 2 04 +. O2 + H 2 0 = NaHS0 4 + NaHSO a•
A suitable solution for general use is prepared by dissolving
50 g. of commercial sodium hyposulphitt! i'n 250-ml. ot water,
30 g. of sodium hydroxide in 40 m!. of water, and mixing the
solutions. * One ml. of this solution will absorb 10·5 ml. of
oxygen. The absorption of oxygen is siower than with alka-
line pyrogallol, but .absorption also takes· place at low tem~
peratures at which the latter reagent is ineffective. Further-
more, no carbon monoxide is evolved or absorbed, and the
reagent is unaffected by unsatuiateg hydrocarbons.
The rate of absorption can be considerably increased by the
audition of sodium anthraquinone-fl-sutphonate. Fieser
(1924) recommends a solution of 16 g. of sodium hypo-
sulphit~, 6·6 g. of sodium hydr,?xide, 2 g. of sodium anthiaqui-
none-fl-sulphonate and 100 m!. of water. When this solution
is exhausted, its colour changes from blood-red to brown;
vigorous shaking is unnecessary. The reagent should nbt be
kept for more ~han one week.
Solid phosphorus. Yellow phosphorus in the form of thin
* For use in the Bunte burette, a solution of 10 g: of Na.S.O. in 50 mi. of
water mixed with 50 m!. of 10 per cent sodium hydroxide solution is preferable.
Gas Analysis 765
sticks may· be employed for the absorption of oxygen in the
Hempel pipette for solid reagents (Fig.. 6-11). and in Orsat
pipettes. A great advantage is that once a pipette has been
charged with the reagent, it does not require renewing except
after long periods. Absorption generally takes about 3
minutes: the oxygen content of the gas should not exceed
50 per cent, otherwise explosions may occur. During the
absorption white clouds of oxide are produced and dissolve in
the water: the clearing of these is a sign that absorption is
complete. "The water in which the phosphorus is immersed
should be changed from time to time in order to remov~ the
phosphoric and phosphorous acids which are formed. The
phosphorus should be shielded from light, otherwise a layer
of red phosphorus will be formed on the surface. The
temperature should be above 15°C-below this temperature
absorption is extremely slow. Even small quantities of
acetylene and unsaturated hydrocarbons inust be absent for
these inhibit the reaction.
D. Carbon monoxide. The usual reagen.t for the absorpt~on
of carbon monoxide is ammoniacal ~uprous chloride solution,
which forms a complex CuCI,2H 2 0,CO. The reagent may be
prepared for use in the ,Ambler or Bone-Wheeler type of
apparatus by adding 75 g. of white cuprous chloride to a
solution of 15 g. of ammonium chloride in 80 ml. of water, and
then adding sufficient concentrated ammonia solution until
solution is complete. The soluti~n should be kept in contact
with copper wire or turnings in a stoppered bottle. For use in
Hempel pipettes or in the Orsat type of apparatus, 125 ml.
of concentrated hydrochloric acid are added to 40 g. of cup-
rous chloride in a suitable flask, followed by the cautious
addition of 130 ml. of concentrated ammonia solution, the
flask being cooled to atmospheric temperature: a little more
ammonia solution (ca. 15 ml.) is then introduced until the
white precipitate just dissolves.
The absorption of carbon monoxide by either of the above
solutions is slow, and intimate contact between the gas and
liquid is necessary: If the gaseous mixture contains more
than 10 per cent of carbon monoxide, it should be treated a
second tilJle with a fresh solution. After the absorption
ammonia must be removed from the gas by washing with
water or with dilute sulphuric acid. Oxygen, acetylene and
to some extent, ethylenes are absorbed by ~he reagent, and
.must previously be removed as already described.
A better reagent is cnprous SUIphate-,B-naphthol, with which
a stable compound Cu 2SO,,2CO is formed. The reagent is
prepared by cautiously adding 200 ml. of concentrated sul-
phuric acid, specific gravity 1·84, to 25 ml. of water; the cool
solution is added a little at a time to 20 g. of specially prepared
cuprous oxide contained in a mortar, and the mixture is w~ll
ground to obtain the solid in fine suspension. 25 Grams of
pure ,B-naphthol is added and incorpora,ted in the mixture by
grinding; the final mixture is filtered through glass wool or
through a sintered glass funnel, and at once transferred to the
absorption pipette. The reagent must be maintained above
15°C and protected frQlIl the atmosphere; the ,B-naphthol
separates out at a lower temperature, and the mixture slowly
oxidises if exposed to the atmosphere.
The cuprous oxide is p1'epared as follows. 100 Grams of cupric acetate are
added to 1 litre of distilled water contained in a 2-litre flask. and the mixture
heated until the acetate is dissolved. The solution is filtered. heated to the
boiling point, treated with a solution of 60 g. of glucose in 400 m!. of water, the
boiling continUed until the blue colour fades somewhat. and the mixture
allowed to stand until the red cuprous oxide separates out. The cuprous
oxide is washed sev.era.l times with water by decantation, transferred to a.
filter, washed once with.alcohol, a.nd finally dried in a vacuum desiccator. •
The cuprous sulphate:',B-naphthol reagent absorbs carbon
monoxid"e completely (1 ml. absorbs ca. 5 m!. of CO), but not
hydrogen or methane; it is, however, rather slow in action
and absorption may take up to 30 minutes. The gas should
be washed with potassium hydroxide solution after contact
with the reagent, since traces of acid. fumes are liable to be
produced.
E. Hydrogen. Hydrpgen is usually deteqnined, together
with methane} etc., by combustion (Section VI, 8). How-
ever, in some cases i.t may be desirable to determine the gas
by absorption. .
The chief'absorbent is palladium. If hydrogen is brought
into contact with palladium in the presence of oxygen, the
hydrogen is removed partly by absorption and partly by
catalytic - combustion. Two-thirds of the contraction is
equal to the volume of hydrogen present. The method is far
from accurate since any carbon monoxide and methane
present may be partly oxidised too. If palladinised asbestos
at 100°C is used, for example, in the Orsat apparat!ls (Section
VI, 60). the percentage of carbon monoxide and methane
burnt will be small, and an approximate value for the hydrogen
content may be obtained; in technical analysis this value is
corrected by determining the carbon dioxide simultaneously:
Gas Analysis 767
produced and adding this value to the initial contraction
observed-a procedure which is open to many objections.
The most satisfactory results are obtainable with a fractional
combustion " pipette" (see Section VI, 60).
VI, 8. Determination by explosion and combustion methods.
-The most important gases determined by combustion
methods are hydrogen, carbon monoxide, and saturated
hydrocarbons (methane CH 4 , ethane CzHe • etc.). The com-
bustion of these gases with excess of oxygen supplies three
results, from which it is possible to determine the amount of
gas blJrned and, in simple cases, its identity. The observa-
tions which can be made are:
{i) Contraction in volume, VI'
(ii) Volume of carbon dioxide produced, V 2 (determined by
absorption in alkali after the comQustion).
(iii) Volume of oxygen consumed, ·v 3 (determined by ab-
sorption after the removal of the carbon dioxide).
In the following table, the volume changes for a number of
simple gases are shown; these refer to one volume of the
combustible gas.
VI V. v.
Hydrogen. H. 3/2 0 1/2
Carbon monoxide. CO 1/2 1 1/2
Methane. CH. .2 1 2
Ethane. CoHo 5/2 2 7/2
Paraffin-; CnHzn + 2 (n + 3)/2 n (3n+ 1)/2
Ethylene. C,B. 2 2 3
Olefine. CnH2n (n/2) + I n 3n/2
The various equations from which the above table is

derived are obvious and include the following for the com-
monly encountered gases:
2H2 + O2 - 2H 2 0;
2CO + O2 2C0 2 ;
CH 4 + 20 2 - CO 2 + 2H 2 0.
Methods of effecting the combustion. A. Explosion. After
the removal of such gases as are determined by absorption.
excess of air or oxygen is added to the residue and the gas
transferred to the explosion vessel, which is usually fitted
with a platinum wire spark gap; the spark is produced by an
induction coil or by a magneto. The Hempel explosion pipette
is shown in Fig. 6-15. It consists of a thick-walled explosion
I bulb A and the levelling vessel B which are connected by
means of thick-walled rubber tubing (" pressure" tubing)
and securely wired on. Two fine platinum wires are fused
iQ.to the explosion ,Pipette, the ends of the wires being about 2
mm. apart. The pipette js filled with mercury. The gas to
be determined together with excess of air or oxygen is
transferred to the bulb A, the mixture brought approximately
to atmospheric pressure, and the stop-cock D closed. The
rubber tube attached to the extremity of E is next closed
with a pinch-clamp, and the rubb~r tube closed by a suitable
glass rod plug (which acts as a safety valve). The pipette js
shaken to ensure complete mixture of the gases, a safety
screen of thick plate glass (preferably of the "triplex"
variety) placed in front of the pipette, and a few sparks passed
. (or until tl)e gas is seen to exoplode). The tap D is immediately-
opened, the gas in the ~ipette transferred without delay to the
burette, and is measured at once if the burette is filled with
mercury or after 2 minutes if the burette contains water. The
technique for explosion in a Bone-Wheeler or similar appara-
tus is self-evident from the above description.
The chief source of error or failure in explosion analysis is an
incorrect proportion of the combustible gas in the explosive
mixture. If the proportion is too low, explo~ion does not
take place or is incomplete; if it is too high and nitrogen is·
present, oxides of nitrogen are produced, giving excessive
contr~ctions and erroneous carbon dioxide figures .. For
hydrogen, the hydrogen cOl!tent must be between 10. and 20
per cent; for metl)ane, the corresponding limits are 6·0 and
8·3 per cent; for mixture.; containing these gases, propor-
tionately intermediate figures apply .. The results for carbon
monoxide oy the explosion method (20 to 28 per cent) are
not always satisfactory.
B. Slow combustion. Catalytic combustion with platinum.
In the presence of oxygen, all combustible gases are oxidised
in contact with platinum wire at about 950°C; hydrogen and
carbon monoxide are oxidised at a much lower temperature.
The usual method of carrying out the combustion is by means
of an electrically heated platinum wire. A slow combustion
pipette, due to Dennis (1899), is shoWn in Fig. 6-16. It is
really a Hempel pipette for solid absorbents (Fig. 6-11) into
the neck of which a 2-holed rubber stopper is inserted which
carries-two glass tubes reaching to within 2 cm. of the top of
the pipette. Stout iron (or nichrome) wires, 2·5 cm. in
Gas Analysis 769
diameter, are mounted inside the tube~ and reach nearly to
the top; the upper ends of the wires carry small iron screws
to which a platinum wire spiral is attached, and the current
passing through the spiral is controlled by a rheostat. The
pipette and levelling tube contain mercury. In use, a known
volume (excess) of oxygen is passed into the pipette followed
by a known volume of the gas to be burned. With non-
explosive gas mixtures, a current is passed through the
platinum spiral, which is maintained at the requisite tempera-
ture with the aid of a rheostat, for 1 or 2 minutes. For
hydrogen and carbon monoxide, one minute's treatment at a
red heat is sufficient, but for methane, which is difficult to
. burn, 2 minutes at a bright yellow heat is usually required.
After combustion, the current is turned off, the pipette
allowed to eool, and the gas residue passed back into the
burette and measured. The reader should be able to adapt
the above experimental details for the Ambler and Bone-
Wheeler type of apparatus.
An improved combustion pipette, due to
Weaver and Ledig (1920) and to Shepherd
(1931), is shown in Fig. 6-26. It is made
of Pyrex glass, has a volume of approxi-
mately 150 mI., and is used with mercury
as the confining liquid; the lower end is
attached by m~ans of rubber tubing to a
levelling bulb. This is incorporated in
Shepherd's •modification of the Orsat
apparatus (see Section VI, 6C). About
20 cm. of wire (90% Pt and 10% Ir alloy)
of 0·16 to 0·17 cm. diameter is employed
in the construction of the coil; the side
arms, through which the wire passes, are
sealed with DeKhotinsky cement.
For mixtures which are liable to ex-
plode, and also to reduce the danger of
explosion in any combustion process,
the combustion pipette of Ambler (1931)- To rubber tubmg .'"'
mercury reservoir
Fig. 6-27-is a valuable improvement
FIG. 6-26.
UpOll the ordinary design. Oxygen is
bubbled through the lower capillary tube side limb, thus
enabling the rate of inflow to be seen and readily adjusted;
in this way the gaseous mixture, already present in the pip-
ette, may be burned in a controlled stream of oxygen.
cc
Combostion with copper oxide. Copper oxide at 270° to
295°C oxidises carbon monoxide and hydrogen rapidly and
completely, methane and other
h ydroGarbons being unaffected.
This. is an example of frac-
tional combustion and' 'is
utilised for separating these
gases. The oxygen is supplied
by the copper oxide; if oxygen
is present, however, it acts as a
catalyst and is not reduced.*
Fractional combustion cannot
be useq.in the presence of un-
saturated hydrocarbons.
VI, 9. Exercises in gas analy..
sis.-The exact nature of these
will depend UP0l,1 the appara-
tus and gaseous mixtures which
are available, upon the par-
FIG. 6-27. ticular interests of the teacher
(and of the student), and upon local conditions. The follow-
ing are suggested as suitable for classes of moderate size.:
they are likely to be useful in alliaborvttories as an introduc-
tion to gas analysis.
1. Determination oj oxygen in air. A Winckler-Hempel
burette, preferably filled with mercury and surrounded by a
water jacket, and a Hempel pipette, filled with phosphorus or
with sodium hyposulphite solution, may be used.
2. Analysis of coal gas by' means of the Bunte burette. Deter-
mination of carbon dioxide with potassium hydroxide solu-
tion~ heavy hydrocarbons with bromine water, oxygen with
sodium hyposulphite solution, and carbori monoxide with
ammc;miacal cuprous chloride solution (for order of absorption, e
see Section VI, 7).
3. Analysis of coal gas by means of tlte Orsat apparatus.
Absorbents as under 2.
4. Analysis of coal gas with the Ambler portable apparatus.
Absorbents as under 2.
• For further details, see Lunge-Ambler, Technical Gas Analysis, 1934. p.
141. The method is widely employed for the determination of hydrogen, and
also of carbon monoxide if saturated hydrocarbons are absent.
GAS-VOLUMETRIC METHODS OF ANALYSIS FOR
SOLIDS AND LIQUIDS
The term " gas-volumetric analysis" is applied to those
processes in which a liquid or solid substance is determined by
utilising some reaction in which it takes place with the
quantitative liberation of
a gas. Such processe~
are not, strictly speaking,
gas analysis, but since 0
they involve technique
and principles which are
commonly encountered in
a
gas analysis, few of the
more important methods
of this class will be des-
cribed.
VI, 10. The Lunge
nitrometer.-This appar-
atus was introduced by
G. Lunge in 1885 for the
analysis of qitrous vitriol
(crude sulphuric acid with A c
dissolved oxides of nitro-
gen from the Chamber
Process). but is of wide
,application in gas-volu-
metric analysis. A dia-
grammatic representation
is given in Fig. 6-28 (a).
It consists of a 50 m!. Qur-
ette A, fitted at the top
with a 3-way tap and
cup-shaped funnel B, and
communicating at the
bottom, by means of (6)
thick-walled rubber tub-
ing, with a levelling tube (0)
C ofthe same diameter. To FIG. 6-28.
increase the accuracy, the volume may be increased by ex-
panding the upper part of A into a 50-100 ml. bulb, only the
lower part being graduated (see B in Fig. 6-7).
771
772 Quantitative Inorganic AnaZvsis
For the analysis of nitrates, nitrites, and nitrous vitriol, the
simple apparatus shown in Fig. 6-28 (a) is employed. For
other gas-volumetric determinations, a reaction vessel, such
as is shown (enlarged) in Fig. 6-28 (b), is attached at D by
means of a short length of " pressure" tubing, secured with
copper wire ligatures; the inner tube of the' reaction vessel is
generally fused to the bottom, and the vessel is closed with a
one-holed rubber bung. In these determinations it will, of
course, be necessary to correct the final volume of the gas to
N.T.P.; this may be avoided by the use of the Lunge gas
volumeter, which has been described in Section VI, 4. The
apparatus is filled wIth mercury, and should always stand on a
" mercury" tray.
VI,ll. Determination of nitrates.-Discussion, If a solution
of a nitrate (or a nitrite) is slJaken with mercury and an excess
of concentrated sulphuric acid, all the nitrogen is set free as
as nitric oxide:
2HN0 3 + 3H 2S0 4 + GHg = 2NO + 4H 20 + 3Hg2S0 4 •
The weight of the nitrate can be computed from the volume
the nitric oxide.
Procedure. For practice in this determination, the student
may determine the percentage of N0 3 in A.R. potassium
nitrate. Sec that both arms of the nitrometer are completely
filied with mercury. Weigh out accurately about 0·12 g.
of potassium nitrate into the cup of the nitrometer. Dissolve
in about 0·5 ml. of water, and draw the solution into the
nitrometer by lowering'the levelling tube and carefully open-
ing the stop-cock. No air must be allowed to enter. Wash
round the cup with 3 or 4 drops of water, and introduce ,the
latter also. Then. place 20 ml. of concentrated sulphuric acid
in the cup, and allow this to enter. Shake the contents of
the nitrometer vigorously for a few minutes, when the reaction
will take place and nitric oxiae will be evolved. Continue the
shaking until no more gas is liberated. The measurement of
the volum~ of the nitric oxide may be effected in either of two
ways; method (b) is the more accurate.
Method (a). Allow the apparatus to stand for at least an
hour so as to acquire the room temperature. It is now neces-
sary to measure the volume of the gas at atmospheric pres-
sure. Adjust the height of the levelling tube to be so much
higher than that in the graduated limb as is necessary to
compensate for the volume of acid in the latter: 1 mm. of
mercury is equivalent to 6·5 mm. of acid. When the adjust-
Gas Analysis 773
ment is made, clamp the tube in the appropriate position.
Test the correctness of the adjustment by placing a few drops
of liquid in the nitrometer cup; upon very cautiously opening
the tap, the liquid should be neither drawn in nor forced out.
Care must be taken not to watm the gas by handling the
measuring tube. Read the volume of the gas, the temperature
and the barometric pressure, and correct the volume to N.T.P.
Method (b). Transfer the nitric oxide to the graduated tube
(filled with ~ercury) of a Lunge gas volumeter (Fig. -6-7)* as
follows. Connect. the nitrometer to the volumeter by means
of as short a piece of rubber tubjng as possible, joining the two
side tubes, and secure with wire ligatures. Displace the air
in the·rubber tube by suitably manipulating the 3-way taps,
and raising the levelling tube of the volumeter. Drive most
of the gas from the nitrometer into the ~volumeter until the
upper level of the acid in the nitrometer is in the constricted
portion just below the tap. Close both taps and sha~e the
nitrometer once :more. When the acid level is in the above
position, It is possibl~ to observe the evolution of a further
small volume of the gas should such still be taking pltlce. If
all reaction has ceased; place a few mL of concentrated sul-
phuric acid in the cup of the volumeter and allow 1 or 2
drops to enter; this ensures that t)1e gas is dry when measured.
Then open both taps and drive the remainder of the nitric
oxide from the nitrometer into the side tube. As soon as the
level of the acid rises to the tap, turn the latter so as to
communicate with the cup of the nitrometer and force all the
acid into the cup. When the mercury below the acid has
risen to the tap, turn the latter back and continue to drive the
gas in the side tube into the volumeter by means of the mer-
cury;leaving the acid in the cup of the nitrometer. In this
way the acid does not come into contact with the rubber
connexion, which it would attack, thus liberating further
qua'ntities of gas. Read the volume of the gas at atmospheric
pressure, and also the temperature. No correction need be
made for water vapour pressure, since the reaction takes place
over concentrated sulphuric acid. Reduce the volume to
N.T.P., and from your result calculate the percentage of
N0 3 in KNO a:
Assume 1 m!. of NO == 0'000,625 g. N2 :0::: 0'002768 g. NOa.t
* For a discussion of the errors in this determination, see Lunge-Ambler
Technical Gas Analysis, 1934, p. 373,
t The control or compensating tube should not be used in this experiment
in order to avoid contamination of the mercury in it.
Note. Two or three consecutive determinations may be carried
out Without washing the mercury. After each series of determina-
tions, the mercury should be emptied out from both nitrometer and
volQmeter, and placed in the Mercury Residues bottle. The whole
apparatus should then be washed out first with .nitric acid, then
with water, and finally with a little acetone, and dried by passing
a stream of air through it. The impure mercury should be washed
thoroughly with water and filtered (see Section. VI, 3)...
.VI, 12. Determination of hydrogen peroxide.-Discussion.
This is an example of a determination by means of a nitro-
meter (or v<;>lumeter) with a special decomposition vessel.
When hydrogen peroxide solution is treated with excess
of acidified potassium permanganate solution, oxygen is
~volved :
2KMnO. -+- 3H 2oO, + 5H2 0 a
= K 2SO, + 2MnSO, + 8H 20 + 50 2 ,
It is evident that half of the oxygen evolved is derived from
the hydrogen peroxide and the other half from the potassium
permanganate. Conversely, by using hydrogen peroxide
~olution' in excess and a measured volume Qf potassium per-
manganate solution, the gas-volumetric method may be
employed for the standardisation of permanganlLte solutions.
The reaction is applicable ~o other oxidising agents, such as
bleaching powder, manganese dioxide, lead dioxide, chromates,
etc. If any of these substances, in solutioI} or in suspension,
and hydrogen peroxide are mixed in the presence of sulphuric
acid, the reagent which is not in excess gives up the whole
·of its available oxygen, and this is increased by an equal
quantity of oxygen from the other reagent:
XO + H 2 0 2 = X + H 2 0 + 02'
The oxygen so produced is measured directly as gas. Several
examples will be given in the following Sections.
Procedure. Dilute 10 ml. oJ "20-volume" hydrogen
peroxide to 40 ml. with distilled water. Measure out accur-
ately 4 ml. of this diluted solution from a burette into the
outer portion of the reaction vessel, Fig. 6-28 (b), and' acidify
with an.equal volume of dilute sulphuric acid. Fill the inner
vessel with a concentrated solution of potassium perrnan-
ganate. The decomposition vessel is connected either to a
simple Lunge nitrometer (Fig. 6-28 (a) ) or to a Lunge gas,
volumeter (Fig. 6-7): in either case the mea>suring tube is
fiJled with mercury. Mix the two liquids taking care to hold
the decomposition bottle so that its contents will not be
Gas Analysis 775
warmed by the heat of the hand, and shake for 1 minute .
•Whilst gas is evolved, the pressure should be reduced in the
measuring tube by lowering the levelling tube. After the
decomposition is complete (excess of permanganate must be
present), place the bottle in a beaker of water and allow it to
attain the temperature of the room. The subsequent pro-
cedure will depend as to whether the simple nitrometer or the
volumeter is used.
With the Lunge nitrometer, adjust the mercury levels, and
read the volume of the oxygen. 'Note the temperature and
barometric pressure, and reduce the volume to N.T.P.
With the Lunge gas volumeter (see Section VI, 4), adjust
the pressure to atmospheric, and isolate the gas in the burette.
As the conditions of the gas in the decomposition vessel are
novy the same as they were at the commencement of the
experiment, the volume of the gas in the burette is equal to
the volume of gas liberated during the'reaction. To reduce
this volume to N.T.P., the control or compensating tube is
employed. If a certain pressure is applied to the gas in the
control tube so that its volume is reduced to 100 ml., it has
been reduced to its volume at N.T.P. If the same pressure
be applied to any other volume of gas under the same con-
ditions as the gas in the control tube, it will also be reduced to
N.T.P. (see Section VI, 4). The actual adjustment is carried
out as follows. Keeping the menisci of the mercury in the
burette and control tube level, raise the levelling tube until
the meniscus in the control tube is exactly at the 100 ml.
mark. The reading of the meniscus in the burette then gives
the volume of gas reduced to N.T.P. ..
From the volume thus obtained, calculate the strength of
the hydrogen peroxide solution in volumes of available oxygen
per unit volume of the original soll;ltion, and also its concen-
tration.
1 Ml. of O2 at N.T.P. ;:;: 0·00151~ g. H 20!.
VI, 13. Evaluation of pvo1usite.-Discussion. Manganese
dioxide reacts with hydrogen peroxide in the presence of
dilute sulphuric acid in accorq.ance with the follbwing equa-
tion ;.
Mn02 + H 20, + H 2 SO. = MnSO. + O2 + 2H2 0.
1 Ml. of O2 at N.T.P. == 0·003885 g. MnO!.
Pyrolusite may contain carbonate; the decomposition
flask may therefore be left unstoppered for about 5 minutes
or llntil &oIl the c&orbonate is decomposed,
776 Quanit'tative Inorganic Analysis
Procedure. Weigh out accurately about 0·2 g. of very finely
ground pyrolusite* and introduce together with abou't 10 m!.
of dilute sulphuric acid into the outer compartment of the'
decomposition vessel (Fig. 6-28 (b)). Place 15 to 25 ml. of
II10-volume "hydrogen peroxid.e into the inner tube. When
no more gas is evolved from the outer compartment, connect
the flask to the gas burette in the usua~ manner, and bring
the hydrogen peroxide solution into contact with the pyro-
lusite. Shake the flask for 2 minutes, at the end of which
time the decomposition should be ~omplete. No black
particles of pyrolusite should remain in the flask. Measure
the volume of oxygen evolved, and from the volume at
N .T.P. calculate the Mn0 2 content of the mineral.
VI,14. Determination of the available chlorine in bleaching
powder.-Discussion. The reaction with acidified hydrogen
peroxide is :
Ca(OCI)2 + 2H 20 2 = CaCl 2 + 2H 20 + ::l02'
i:e., 1 Ml...of O2 at N.T.P. == 0·003167 g. of available chlorine.
Procedure. Place 25 ml. of the well-mixed bleaching powder
suspension (Section m, 111, Procedure A) into the outer
compartment and 6 m!. of approximately 5-volume "II
hydrogen peroxide into the inner tube. Carry out the deter-
mination in the usual manner. Calculate the percentage of
available chlorine in the sample of bleaching powder.
VI, 15. Evaluation of zinc dust.-Zinc is ~requently used as
a reducing agent, and its value for this purpose may be
determined by measuring the volume of hydrogen evolved-
when a weighed sample is treated with excess of dilute acid.
Weigh out accurat~ry about 0·1 g. of the zinc dust into th~
outer compartment, and place 25 m!. of 1 : 3 sulphuric acid
in the inner tube .• Carry out the determination in the usual
manner. From the volume of hydrogen evolved, reduced to
N.T.P., calculate the percentage of zinc in the sample 'of zinc
dust.
VI, 16. Determination of ammonium salts.-Discu.ssion.
Ammonium salts are decomposed by sodium hypobromite
solution with the evolution of nitrogen:
2NH3 + 3NaOBr = 3NaBr + 3H 2 0 + N 2 •
'" Ridsdale's .. Pyrolusite, No. 18" (one of the Analysed Samples for Stu-
dents) is suitable for this determination.
Gas A lla.lysis 777
It is found, however, that only 97·5 per cent of the nitrogen is
evolved.. For practice in thi~ determination, the student may
either assume this factor and use it to determine the per-
centage of NH3 in an ammonium salt (e.g., ammonium
chloride) or he may employ a pure ammonium'salt (e.g., A.R.
ammonium chloridej and redetermine this factor.
Urea reacts similarly:
CO(NH 2)2 + 3NaOBr = CO 2 + 3NaBr + 2H 2 0 + N.2'
but here only 91·5 per cent. of the nitrogen is evolved (the
carbon dioxide is absorbed by the excess of sodium hydroxide
in the sodium hypobromite solution). Similar exercises may
be carried out with urea.
Procedure. Weigh out accurately about 0·25 g. of the
ammonium salt (or about 0·12 g. of urea), dissolve in about 20
mI. of water, and place in the outer portion of the decomposi-
tion vessel (Fig. 6-28 (b)). Prepare a sodium l:typobromite
solution by dissolving 16 g.' of sodium hydroxide in 40 m!. of
water and add 4·0 ml. of bromine, measured in a small
graduated cylinder. Fill the inner tube with the freshly-
prepared sodium hypobromite solution. Connect up the
apparatus to a burette filled with mercury- as in Section
VI, 11; shake the vessel gently to complete the evolution of
nitrogen. Determine 'the volume of nitrogen at N.T.P.
Use this result either to determine the purity of the initial
s~bstance, assumIng th.e factors given. in the discussion
(1 Ml. of N z at N.T.P. == 0·001559 g. NHa ;:;;: 0·002929 g. urea)
or determine the factor on the assumption that the initial
compound (ammonium salt or urea) is,pure.
APPENDIX
A, 1. INTERNATIONAL ATOMIC WEIGHTS, 1939
El~ment ~ Symbol At. Wt. Element Symbol At. Wt.

Aluminium ..... Al 26·97 Neon ........... Ne '20'183
Antimony , ..... Sb 121·76 Nickel .......... Ni 58·~9
Argon .....•.... A 39·944 Niobium
Arsenic ...... : .. As 74·91 (Columbium) Nb (eb) 92·91
Barium ......... Ba 137 ·36 Nitrogen ....... N 14·008
Bervllium ...... Be . 9'02 Osmium ....... Os 190·2
Bismuth ........ Bi 209'00 Oxygen ........ 0 16·0000
Boron .......... B 10'82 Palladium ...... Pd 106·7
Bromine ..... : .. Br 79·916 Phosphorus ..... P 30·98
Cadmium ....... Cd 112·41 Platinum ....... Pt 195·23
Caesium .~ ..... Cs 132'91 Potassium ....... K 39·096
Calcium ......... Ca 40·08 PraseodymIum ... Pr 140·92
Carbon ......... C 12·010 Prot6actinium ... Pa 231
Cerium ......... Ce 140'13 Radium ......... Ra 226·05
Chlorine ....... Cl 35·457 Radon ......... Rn 222
Chromium ...... Cr 52·01 Rhenium ........ Re 186·31
Cobalt ...... , .. Co 58·94 Rhodium ....... Rh 102·91
Copper ......... eu 63'57 Rubidium ...... Rb 85·46
Dysprosium ..... Dy 162·46 Ruthenium ..... Ru 101·7
Erbium ........ Er 167·2 Samarium ...... Sm J50'43
Europium ...... Eu 152'0 Scandium ..... Sc 45·10
Fluorine ...•.... F. 19·00 Selenium ....... Se 7-8·96
Gadolinium ..... Gd 156·9 Silicon ......... Si 28·0.11
Gallium •....... Ga 69·72 Silver .......•... Ag 107·880
Germanium ..... Ge 72·60 Sodium ......... Na 22·997
Gold ........... Au 197·2 Strontium ...... Sr 87·63
Hafnium ....... Hf 178·6 Sulphur ........ S 32·06
Helium ......... He 4'003 Tantalum ....... Ta 180·88
Holmium ...... Ho 163·5 Tellurium ....... 'Ie 127·61
Hydrogen ...... H l'OO~1 Terbium ........ Tb 159·2
Indium ......... In 114·76 . Thallium ....... Tl 204·39
Iodine .......... 1 126·92 Thorium ........ Th 232·12
Iridium ........ Ir 193·1 Thulium ........ Tm 169·4
Iron .......... Fe 55·84 Tin ............. Sn 118·70
Krypton ........ Kr 83·7 Titanium ....... Ti 47·90
Lanthanum ..... La 138'92 Tungsten ..•.... W 183·92
Lead ........... Pb 207·21 Uranium ....... U 238·07
Lithium ........ Li 6'940 Vanadium ....... V 50·95
Lutecium ....... Lu 175·0 Xenon ......... Xe 131·3
Magnesium ..... Mg 24·32 Ytterbium ....... Y d 173·04
Manganese ..•... Mn 54·93 Yttrium ........ Y 88·92
Mercury .•..••. Hg 200'61 Zinc ......••.. :Zn 65·38
Molybdenum .... Mo 95'95 Zirconium ....•. Zr 91·22
Neodymium '" ;Nd 144·27
778
Appendix 779 _
A.2. CHEMICAL FACTORS·
Sought Found Factor Log Factor
Pb PbSO, 0·68325 1·83458
PbCrO, 0·64108 1·80691
PbMoO, 0·56436 }·75156
. PbO! 0·86623 }·93763
Ag AgCI 0·75263 }·87658
Hg HgS 0·86221 }·93561

Hg[Zn(CNS)J 0·40259 1·60486
[Cu (C,H lGN,)] [HgIJ 0·22488 }·35195
Bi BiOI 0·59388- }'77370

Bi(C8H sOa) 0·62937 1·79890
BiPa 0·89700 1·95279
Bi(Cg H80N)s 0·32583
. }'51299
Cd CdMoO, 0·41273 }'61566

[(C18HgN)2Ha] CdI, 0'1l464 1·05934
Cd(C 1oHe02N)z 0·24612 Y·39115
[Cd(CoH5N)z] (CNS)! - ~'29065 Y·46336
.
Cu Cu 2'(CNS)! 0·52257 1·71814
CU(C14HllOsN) 0'22011 1·34264
Cu(C,HeOaN)! 0·18930 1·27715
Cu(C1oHeOzN)s,HP 0·14926 !'1739~
[CU(C5H 5N)a] (CNS)2 0·18812 1·27444
[CU(C,H 18N ,)][HgIJ 0·07126 2·85285
As AszSa 0·60903 1'78463

Mg(NH,)AsO,,6HzO 0·25887 1·41309
Mg,AssO, 0·48258 1·68356
VaOs 0·26683 1·42624
Sb SbsS s 0·71687 1·85544

SbaO, 0·79189 1·89866
Sb(CeHoO,) 0·46321 1·66578
Sn Sn02 0·78766 1·89634
Mo PbMoO" 0·26133 1·41719

MoOs 0·66655 1'82383
Mo02 (C oH.ON)z 0·23051 1·36269
Te , TeO. 0·79951 }·90282

-For discussion on chemical factors, see Section IV, 3.
CHEMICAL FACTORS-Continued
Sought - Found Factor Log. Factor
Pd Pd(C,HP2N 2l2 0·31669 1·50064
Pd(C loH e0 2N)2 0·23658 1·37398
Pd(C 1O H 6 0 3N)2 0·22091 1·34421
Al IA1AI(G
20 3
g H 6ON)s
0·52913
0·05871
1·72357
2·76868
-
Be _BeO 0·36051 1·55692
'Cr CrS0 3. 0·68425 1·83522

- BaCrO, 0·20527 1·31233
-
Fe I- Fe20 a 0·69940 1·84472
Ni Ni(C4 HP2N 2)2 0·20314 1·30780

Ni(C7 H eOsN )2 0·1'7734 1·24881
[Ni(CsHsN),J(CNS)z "0·11947 1·07727
Ni(C 14H n 0 2N 2)z 0·10926 1·03844
I
I Co Co(CloHe02N)3,2HaO 0·09640 2·98406
Co(CloHeOsN) s . 0·09454 2·97564
[CO(C5HSN)J (CNS)2 0·11992 1·07890
Zn Zn(NH,)PO, 0·36648 1·56405

Zn 2P a07 0-42912 1·63257
ZnS 0·67098 1·82671
Zn(CloH602N)2,2H20 0·14668 1·16638
[Zn(CsHsN)J(CNS)s 0·19245 1·28431
Mn I Mn::P207 0"'38708 1·58780
U UsDs 0·84802 1·92841

UO a UsOs 0·96200 1·98318
U UO s(C 9 H 6 ON)2,C gHPN . 0·33840 1·52943
Th Th0 2 0·87884 1·94391
Ce Ce0 2, 0·81409 1·91067
Ti I TiO
TiO(C H eON)2
g 0·13601
0·59950
1·13356
1·77779
a
Zr I Zr02
Zr P 207
2 0·51190
0·74030
1·70918
1·86941
Appendix 781
CHEMICAL FACTORS-Continued
Sought Found Factor Log. Facto,!
TI TI 2CrO,
T1 3 [Co(N0 2).]
0·77894
0·64669
1·89150
1·81070 I
Ca CaC2O"H 2O 0·27430 1·43822
CaC0 3 0·40044 1·60254
CaO 0·71469 1·85412-
Sr SrSO,
SrO . 0·47705
0·R4.'560
1·67857
1·92717
Ba BaSO, 0'58S47 1·76972

BaCrO, 0·54213 }·73411 \
Mg Mg(C,H60N)2,2H zO 0·06976 . 2·84359 I

Mg(C.H.ON)2
Mg{NH 4)PO,,6H zO
0·07780
0'09909
2·89096
2·99602 I
Mg ZP,.07
.
0·218fil 1·33947
I
Na Na:SO, 0·32376 1·51022
NaZn(UOZ)3(C zH 30 2),,6H,.G 0·01495 2·17475
NaMg(UOZ)3(CzHa02),,6'5HtO 0·01274 2·10500
K K,[PtCIJ 0·16083 1·20637

KCI0, 0·28217 1·45052
KzSO, 0·44888 1·65213
KzN a[Co(N °
2 ).] H 2O
--
0·17215 1·23591
Li Li 2SO, O·1262fl 1·10123

-LiCI 0·16369 1·21402
NH, (NH,MPtCl.J O'OS12fl 2·90983
W W0 3 0·79303 1·89929
W0 2(C,H 60N)2 ~ 0·36477 1·56202
,
-
-Anions
(.1 AgCl 0·24737 1·39334
Br AgBr
I 0·42555 1·62895
I AgI 0·54055 1·73283

PdI 2 0·70405 1·84761
i
782 Quantitative ltlOrganh Analysis
CHEMICAL FACTORS-Contin,1ted
Sought Found Factor Log Factor

CNS AgCNS 0·34996 1·54402
Cu 2 (CNS)z 0·47744 1·67891
BaSO, 0·24882 }'39589
CN AgCN 0'19413 1'28850
F PbCIF 0'072(}I 2·86100

. CaFs 0·48668 1·68724
(C 6 H.)3SnF 0·05149 2·71173
ClO a AgCl 0·58224 }'76510
ClO, AgCI ()·69387 1·84128

KCIO, 0·71783 1·85602
SO, BaSO, 0·41153 }'6144O
5 BaSO, 0'1~735 1·13783
SOs BaSO, 0·34299 }'53528
S20 a BaSO, 0·24017· 1·38052 •
PO, MgZP Z0 7 0·85337 1·93114

P ZO&,24MoO a 0·05281, 2·72275
HPO a HgsClz 0·16942 }·22896

MgZP ZP 7 0·71869 1·85654
RzPO z RgzClz 0-06884 '2-83781

Mg ZP Z0 7 0·58401 1·76642
Cp, CaCp"HsO
, 0·60238 1·77987
CaCO a 0·87941 }'944:19
CaO 1·56954 0·19577
Si0 3 SiOs 1·26640 0·10257
SiF6 SiOz 2'36530 0'37389
NO a COOH 16N,.HNO a 0·16519 }·21799

Appendix 783
At 3. THE LITERATURE OF ANALYTICAL CHEMISTRY
The following list is not intended to be exhaustive-; it indicates books
and journals which may.well form the nu~leus of a library for analytical
chemistry. A number of comparatively elementary works on general
analysis are included because these contain references to the scientific
literature. .
1. Theory of Analytical Chemistry.
H_. Bassett. The :r~eory of Quantitative A iialysis. G. Routledge, i925.
T. B. Smith. Analytical Processes. E. Arnold, 1929.
2. General Reference Works.

W. F. Hillebrand and G. E. F. Lundell. Applied Inorganic Analysis.
J. Wiley; Chapman and Hall, 1929.
W. W. Scott and N. H. Furman. Standard Methods of Chemical
. Analysis. Vol. I. The Elements. 5th Edition. Van Nostrand;
Technical Press, 1939.
J. W. Mellor and H. V. Thompson. A Treatise on Quantitative Inorganic
Analysis. 2nd Edition. C. Griffin and Co., 1938.
G. E. F. Lundell and J. I. Hofmann. Outline of Methods of Chemical
Analysis. J. Wiley; Chapman and Hall, 1938.
A. D. Mitchell and A. M. Ward. Modern Methods in Quantitative
Chemical Analysis. Longmans, Green and Co., 1932.
C. A. Mitchell. Recent Advances in Analytical Chemistry. Vol. II.
Inorganic. Churchill, 1931. .
3. Text Books.
F. P. Treadwell-W. T. Hall. Analytical Chemistry. Vol. II. Quanti-
tative Analysis. .8th Edition. J. Wiley; Chapman and Hall, 1935.
J. M. Kalthoff and E. B. Sandell. Text Book of Quantitative Inorganic
Analysis. Macmillan, 1936.
H .• H. Willard and N. H. Furman. Elementary Quantitative A Ilalysis.
2nd Edition. Van Nostrand; Macmillan, 1935.
H. A.~Fales. l1zorganic Quantitative Analysis. Century, 1925.
W. D. Treadwell. Tabelle?J und Vorschriften zur Quantitativm Analyse.
Franz Deuticke, 1938.
•
4. Voluihetric Analysis.
F. Sutton-A. D. Mitchell. A Systematic Handbook of Volumetric
Analysis. 12th Edition. Churchill, 1935.
W. Bottger-R. E. Oesper. Newer Methods of Volumetric Analysis.
Van Nostrand; Chapman and Hall, 1938.
1. M. Kolthoff-N. H. Furman. Volumetric Analysis. 2 Vols. J,
Wiley; Chapman and Hall, 1928-29.
G. S. J~mieson.. Volumetric Iodate Methoils. Chemical Catalog. Co.,
1926.
E. Knecht and E. Hibbert. New Reduction Methods in Volumetric
Analysis. 2nd Edition. Longmans, Green and. Co., 1925.
* A more detailed list of books is to be found in E. J. Crane and A. M.
Patterson, A Guide to the Literature of Chemistry (1. Wiley, 1927) and in J.
Olsen, Van Nostrand's Ohemical Atmual (1935).
784 Q1tantitative In_organic Analysis
5. Electro-analysis. E1ectrochemistrJ~
E. F. Smith. Electro-analysis. 6th Edition. Blakiston, 1919.
A. Classen and H. Danneel. Quantitative Analyse durch Elektrolyst
7th Edition. J. Springer, 1927.
A. Fischer-A. Schleicher. Elektro-analytischer Schnellmethoden. 2nd
Edition. F. Enke, 1926.
W. Bottger. Physittalische Methoden del' analytischen Chemie. Band
II. Akad. Verlag, 1936.
H. T. S. Sand. Electrochemistry and Electrochemical Atlalysis. Blackie.
1940
S. Glasstone. The Electrochemistry of Solutions. 2nd Edition. Methuen,
1937.
D. A. MacInnes. The Principles of Electrochemistry. Reinhold Pub-
lishing ~orporation; Chapman and Hall, 1939.
N. A. McKenna. Theoretical Electrochemistry. Macmillan, 1939.
6. Gas Analysis.
L; M. Dennis and M. L. Nichols. Gas Analysis. Revised Edition.
Macmillan, 1929.
G. Lunge-H. R. Ambler. Technical Gas Analysis. Revised Edition.
Gurney and Jackson, 1934. .
G. W. Himus. Fuel Testing. Leonard Hill, 1932.
J. S. Haldane and J. J. Graham. Methods of Air Analysis. 4th Edition
C. Griffin and Co., 1935.
7. Colorimetric Analysis.
J. H. Yoe. Photometric Chemicl)l Analysis. Vol. I. Colorimetry.
Vol. II. Nephelometry. J. Wiley; Chapman and Hall, 1928-29.
E. D. Snell and C. R. Snell. Colorimetric Methods of Analysis. Vol. I.
Inorganic. Van Nostrand; Chapman and Hall, 1936.
A bsolutkolorimetrische M etallanalysen mit dem' P~tlfrich-Photomete',
Carl Zeiss, Jena, 1938. -
S. Electrometric JV[eth~.
I. M. Kolthoff and N. H. Furman. Potentiometric Titrations. 2nd
Edition. J. Wiley; Chapman and Hall, 1931.
H. T. S. Britton. Conductometric Analysis. Chapman and Hall, 1934.
H. T. S. Britton. Hydrogen Ions. 2nd Edition. Chapman and Hall,
1932. ..
W. M. Clark. The Determination of Hydrogen Ions. 3rd Edition.
William and Wilkins; Bailliere, Tindall and Cox, 1928.
1. M. Kolthoff and N. H. Furman, Colorimetric and Potentiometric
Determination of pH. J. Wiley; Chapman and Hall, 1931.
9. Micro-analysis, Organic Reagents, etc.
F. Emich-F. Schneider. Micro-chemical LabC'ratoryManual. J. Wiley;
Chapman and Hall, 1932. _
F. Pregl-H. Roth-E. B. Daw. Quantitative Organic Micro-analysis.
Churchill, 1937.
J. R. Niederl and V. Niederl. Micro-methods oj Quantitative Organic
Elementary Analysis. J. Wiley; Chapman and Hall, 1938.
E. M. Chamot and C. W. Mason. Handbook of Chemical Microscopy.
Vol. 1. 2nd Edition, 1938. Vol. II. 1st Edition, 1930. J. Wiley;
Chapman and Hall.
Appendix 785
F. Feigl-J. W. Mathews. Qualitative Analysis by Spot Tests. Norde-
mann Publishing Co., 1937.
H. V. A. Briscoe and J. W. Mathews. Mic1'o-chemical Methods .suitable
for General A nalytical Practice. Institute of Chemistry, London,
1934.
E. M. Chamot. Chemical Microscopy (in Scott-Furman, Standard
Methods of Chemical Analysis, Vol. II, p. 2435, 1939).
L. T. Hallett. Quantitative Micro-chemical Analysis (in Scott-Furman,·
Standard Methods of Chemical Analysis, Vol. II, p. 2460, 1939).
Organic Reagents for Metals. 3rd. Edition. Hopkin and Williams,
London, 1938.
C. J. van Nieuwenburg, W. BOttger, F. Feigl, A. S. KGmarovskyand
N. Strafford. Tables of Reagents for Inorganic Analysis (Premier
Rapport de la " Commission internationale des Reactions et Reactifs
analytiques nouveaux" de 1'" Union Internationale de Chimie,"
Paris). Akad Verlag. Leipzig; H. K. Lewis, Lo~don, 1938.
10. Qualitative Analysis.

R. K. McAlpine and B."A. Soule. Qualitative Chemical Analysis. Van
Nostrand; Chapman and Hall, 1933.
A. I. Vogel. A Text Book of Qualitative Chemical Analysis. Longmans,
Green and Co., 1937. . ....
F. P. Treadwell-W. T. Hall. Analytical Chemistry. Vol. I. Qualita-
tive Analysis. 9th Edition. ]. Wiley; Chapman and Hall, 1937.
A. A. Noyes anq W. C. Bray. A System of Qualitative Analysis for the
Rarer Elements. Macmillan, 1927.
1,).. Metallurgical Analysis.
G. E. F. Lundell, J. I. Hofmann and H. A. Bright. Chemical Analysis
of Iron and Steel. ]. Wiley; Chapman and Hall, 1931.
W. A. Naish and J. E. Clennell. Select Methods of Metallurgical
Analysis. Chapman and Hall, 1929.
N. W. Lord and D. J. Demorest. Metallurgical Analysis. 5th Edition.
McGraw-Hill, 1924. •
U.~. Steel Corporation. Sampling and Analysis of Carbon and Alloy
Steels. Reinhold Publishing Corporation; Chapman and Hall, 1938.
F. Ibbotson and L. Aitchison. The Analysis of Non-Ferrous Alloys.
2nd Edition. Longmans, Green and Co., 1922.
W. B. Price and R. K. Meade. The Technical Analysis of Brass and
Non-Ferrous Alloys. J. Wiley; Chapman and Hall, 1922.
E. Gregory and W. W. Stevenson. Chemical Analfsis of Metals and
Alloys. Blac~ie, 1937.
12. Ore Analysis.
A. H. Low. Technical Methods of Ore Analysis. 10th Edition. J.
Wiley; Chapman and Hall, 1936.-
H. S. Washington. Chemical Analysis of Rotks. 4th Edition. J.
Wiley; Chapman and Hall, 1930.
W. R. Schoeller and A. R. Powell. The Analysis of Minerals and Ores
of the Rarer Elements. C. Griffin and Co., 1919.
A. W, Groves. Silicate Analysis. T. Murby, 1938.
H. F. Harwood. Analysis of Rocks and Minerals. Annulli Reports on
the Progress of Chemistry, 1938, 34, 464-.480.
13. Water Analysis.
J. C. Thresh, J. F. Beale and E. V ..Suckling. The Examination oj
Watlr and Water Supplies. Churchill, 1933.
W. P. Mason and A. M. Buswell. Examination of Water. 6th Edition.
J. Wiley; Chapman and Hall, 1931.
Standard Methods for the Examination of Water and Sewage. 8th
Edition. American Public Health Association, 1936.
D. K. French. Wate, Analysis (in Scott-Furman, Standard Methods of
Chemical Analysis, Vol. II, p. 2046; 1939).
14. Reagents.
.. Analar .. Standards for Laboratory Chemicals. 2nd Edition. Hopkin
and Williams Ltd., London and British Drug Houses Ltd., Lonpon,
1937.
J. Rosin. Reagent Chemicals and Standards. Van Nostrand; Chap-
man and Hall, 1937. -
Re.ports of the Committee on Guaranteed (or Analytical) Reagents of the
American Chemical Society. Ind. Eng. Chern., 1925,17,756; 1926,
18, 1136, 759; 1927, 19, 645, 1369; 1928, 20,979. Ind. Eng. Chern:
Anal. Edn., 1929, 1, 171; 1930,2, 351 ;19!H, 3, 221; 1932,4, 154,
347; 1933, 5, 289.
15. Cement.
R. K. Meade. Portland Cement. Chemical Publishing Co., 1926.
W. C. Hanna. Cement (in Scott-Furman, Standard Methods of Chemi-
cal Analysis, Vol. II, p. 1597; 1939).
16. General Industrial Analysis.
W. W. Scott-No H. Furman. Staudard Methods of Chemical Analysis,
Vol. II. Special Subjects. 5th Edition. Van Nostrand-Technical
Press, 1939.
G. Lunge-C. A. J(ea1le-P. C. L. Thorne. Technical Methods of Chemical
Analysis. 2nd Edition. 3 Vols. Gurney and Jackson, 1924-31.
A. H. Allen. Commercial Organic Analysis. 5th Edition. 8 Vols .
• Blakiston; Churchill, 1923-30.
'I.7. Agricultural Analysis, Food Analysis, etc.
W. W. Skinner et al. Official and Tentative Methods of Analysis. 4th
Edition. Association of Official A.gricultural Chemists. Washington,
1935.
E. G. Mahin .and R. H. Carr. Quantitative Agricultural Analysis.
McGraw-Hill, 1923.
C. H. Wright. A.cricultural Analysis. T. Murby, 1938.
H. W. Wiley. Principles.and Practice of Agricultural Analysis. Vol. [.
Soils. 3td Edition, 1926. Vol. II. Fertilisers and Insecticides. 3rd
Edition, 1931. Chemical Publishing !=C.
A. G. WOOdman. Food Analysis. 3rd Edition. McGraw-Hill, 1931.
A. W. Blyth-M. W. Blyth-H. E. Cbx. Foods: Their Composition and
Analysis. 7th Edition. C. Griffin and Co., 1927.
H. E. [:.ox. The Chemical Analysis of Foods. 2nd Edition. Chtlrchill,1938.
C. G. Moor-W. Partridge-I. R. Nicholls. Aids to the Analysis of Foods
and Drugs. 5th Edition. Bailliere, Tindall and Cox, 1934.
C. K. Tinckler and H. Masters. ApP.lied Chemistry. Vol. I. Wat"r
Detergents, Fuels, Textiles, etc . . 3rd Edition, 1935. Vol. II. Foods.
2nd Edition, 1932. The, Technical ~,
Appendix 787
18. Apparatus, etc.
Much valuable information concerning apparatus is to be found in the
catalogues of various manufacturers. A selected list is given below.
Some firms publish booklets from time to time in which recent improve-
ments of laboratory apparatus and technique are described.
A. Gallenkamp and Co. Ltd., 17-29 Sun Street, London, E.C.2, England.
Baird and Tatlock (London) Ltd., Cross Stre~t, London, E.C.I, England.
Griffin and Tatlock Ltd., Kemble Street, London, W.C.2, England.
F. E. Becker and Co., Hatton Wall, London, E.C.I, England.
Fisher Scientific Co., 709-719 Forbes Street, Pittsburgh, Pa., U.S.A.
~ E. H. Sargent and Co., 155-165 East Superior Street, Chicago, U.S.A.
Central Scientific Co., 1700 Irving Park Blvd., Chicago, U.S.A.
Eimer·and Amend, 205-223 Third Avenue, New York, N.Y., U.S.A.
l!ulletin and Laboratory Notes, published by Baird& Tatlock (London}Ltd.
The Laboratory, published by the Fisher Scientific Co.
Cenco News Chats, published by the Central Scientific Co.
Twe Nivoc Supplement, published by F. E. Becker and Co.
Short pamphlets announcing new or improved apparat1}s are pub-
lished from tim,e to time by the other firms nofed above.
19. Chemical and Physical Constants.
E. W. Washburn et al. International Critical Tables. 7 Vols. Mc-
Graw-Hill, 1926-30.
H. Landolt-R. Bornstein. Physikalisch-Chemische Tabellen. 5th
Edition. J. Springer, 1923-36.
G. W. C. Kaye and T. H. Laby. Tables of Physical and Chemical
Constants. 8th Edition. Longmans, Green ap.d Co., 1936.
N. A. Lange. Handbook of Chemistry. 2nd Edition. Handbook
Publishers, 1937. .
C. D.-Hodgman. Handbook of Chemistry and Physics. 20th Edition.
Chemical Rubber Publishing Co., 1935.
J. C. Olsen. Van Nostrand's Chemical Annual. 7th Issue. Van
Nostrand; Chapman and Hall, 1935.
E. Hope. The Chemists' Year Book. Sherratt and Hughes, 1938.
20. lournals .
• A list of the most important Journals in which the results of original
research in analytical chemistry are published is given below. Many
memoirs on chemical analysis appear in numerous other scientific
journals; abstracts of these may be found in the British Chemical
Abstracts published by the Bureau of Chemical Abstracts, London, or in
Chemical Abstracts published by the American Chemical Society.
Valuable summaries appear in the Annual Reports on the Progress of
Chemistry published by the Chemical Society, London.
The Analyst.
Industrial and Engineering Chemistry, Analytical Edition.
Annales de Chimie analytique 1!t de Chimie appliquee et Revue de Chimie
analytique reunies. ...
Zeitschrift fur analytischen Chemie.
Journal of the Chemical Society.
Journal of the A merican Chemical Society.
Bulletin de la Societl chi,mique de France.
Berichte der deutschen chemischen GesellsGhaft.
Recucil des Travaux chimiques des Pays-Bas et de la Belgique.
Journal of Research of the National Bureau of Sfandards.
788 Quantitative Inorganic A Italyst's
A, 4. DENSITIES OF ACIDS AT 20nC
(DENSITIES AND PERCENTAGES BY WEIGHT ARE BASED ON WEIGHTS IN VACUO,
AND THE PERCENTAGE BY WEIGHT REFERS TO THE FORMULA GIVEN:')
Density
Per Cent -
By Weight Ha SO• HNO. .H.CzH 3 Oz H 3 PO. Hel
1 1'0051 1·0036 0'9996 1'0038 1·0032

2 1·0118 1·0091 1·0012 1:0092 1·0082
3 1·0184 1·0146 1·0025 - -
4 1·0250 1·0201 1·0040 1·0200 1·0181
5
6
1·0317
1'0385
1·0256
1·0312
1·0055
1·0069
-
1'0309
-
1·0279
7 1·0453 1·0369 1·0083 -
1·0420
-
8 1·0522 1·0427 • 1'0097 1'0376
9 1·0591 1·0485 1·0l11 - -
10 1·0661 1·0543 1·0125 I'P532 1'0474
11· 1'0731 1'0e02 1·0139 - - •
12 1'0802 1·0661 1·0154 1·0647 1·0574
13 1'0874 1·0721 1'0168 - -
14 1·0947 1·0781 1·0182 1·0764, 1·0675
15 1'1020 1·0842 1·0195 - -
16 1'1094 1·0903 1·0209 1·0884 )'0776
17 1·1168 1·0964 1·0223 - -
18 1·1243 l-l026 1'0236 ) ·1008 ) '0878
19 H318 H088 1·0250 - -
20 1'139~ 1·1150 1·0263 1·1134 1·0980
21 1-1471 1-1213 H)276~ - -
22 1·1548 1-1276 1·0288 1·1263 1·108:J
23 1·1626 1·134.0 1·0301 - -
24 1-1704 1-1404 1·0313 1·1395 I·U87
25 1·1783 1·1469 1·0326 - -
26 1'1862 1·1534 1·0338 1·1529 1·1290
27 1·1942 1-1600 1·0349 - -
,
28
29 . 1·2023
1·2104
1,,1666
1·1733
1·0361
1·0372
1·1665
_-
1·18()5
1·1392
-
1·1493
30 1·2185 1·1800 1'0384
:n
32
1·2267
1·2349
1·1867
1·1934
1'0395
1·0406
-
-
-
1·1593
.
33 1·_......" 1·2002 1·0417 - -
34 1:2515 1·2071 1'0428 - 1·1691
35 1·2599 1·2140 1·0438 1·216 -
36 1·2684 1·2205 }·0449 - }·1789
37 1·2769 1·2270 1·0459 - -
38 }·2855 1·2335 1·0469 - H885
39 1·2941 1·2399 1·0479 - -
40 1·3028 1·2463 1·0488 1·254 1-1980
41 1·3116 1'2527. 1·0498 - -
42 , 1·3205 1·2591 1'0507 - -
43 1'3294 1·2655 1·0516 - -
44 1·3384 1·2719 "1·0525 - --
45 1·34'76 1·2783 1·0534 1·293 -
46 1·3569 1·2847 1'0542 - -
47 1·3663 1·2911 1·0551 - -
48 1·3758 1·2975 1,,0559 - -
49 }·3854 1·3040 1·0567 - -
50 1·3951 1·3100 1'0575 1·335 -
Appendix 789
DENSITIFB OF ACIDS AT 20°C-Continued
Density
Per Cent
By Weight
I HaSO,
I HNO.
I H.CaH.Os I H.PO,
51 1·4049 1·3160 1·0582 -

52 1·4148 J.·321p 1·0590 -
53 1·4248 1·3278 1·0597 -
54 1·4350 1·3336 1-0604 -
55 1·4453 1·3393 1·0611 1·379
56 1·4557 1·3449 1·0618 - -
57 1·4662 1·3505 1·0624 -
68 1·4768 1·3560 1·0631 -
59 1·4875 1·3614 1·0637 -
60 1·4983 1·3667 1·0642 1·426
61 1·5091 1·3719 1·0648 -
62 1·5200 1·3769 1·0653 -
63 1·5310 1·3818 1·0658 -
64
65
1·5421
1·5533
1·3866
1·3913
1·0662
1·0666
-
}·475
66 1·5646 1·3959 1·0671 -
67 1·5760 1:4004 1·0675 -
68 1·5874 1·4048 1·0678 -
69 1·5989 1'4091 1·0682 -
70 1·6105 1·4134 1·0685 1·526
71
72
1·6221
1·6338
1·4176
1·4218
1·0687
1'0690 --
73 1·6456 1·4258 1'0693 -
'Z4 1·6574 1·4298 1'0694 ~
75 1·6692 1·4337 1'0696 1·579

76 1·6810 1'.4375 1'0698 -
77 1·6927 1·4413 1'0699 --
78 }·7043 1·4450 }'0700 -
79 1·7158 1·4486 1-0700 -
80 1·7272 1·4521 1'0700 1·633
81 1·7383 1'4555 1,0699 -
82 1·7491 1·4589 1·0698 -
83 1·7594 1·4622 1·0696 -
84 1·7693 1·4655 1'0693 -
85 1·7786 1·4686 1·0689 1·689
86 1·7872 1·4716 1'0685 -
87 }'7951 }·4745 1,0680 --
88 1·8022 1·4773 1·0675 -
89 1·8087 1·4800 1·0668 -
90 1·8144 1·4826 1·0661 1·746
91 1·8195
1·8240
1·4850 }'0652 -
92 1·4873 1·0643 1·770
93 1·8279 1·4892 1·0632 -
94 1·8312 1·4912 1·0619 1·794
95 1·8337 1·4932 1·0605 , -
96 }·8355 1·4952 1,0588 1·819
97 1·8364 1·4974 1·0570 -
98 1·8361 1·5008 1·0549 1·844
99 1·8342 1·5056 1·0524 -
100 1·8305 1'512!T 1·0498 1·870
790 Quantitativ~
Inorganic Analysis
At 5, DENSITIES OF ALKALINE SOLUTIONS AT 20°C
(DENSITIES AND PERCENTAGES BY WEIGHT ARE BASED ON WEIGHTS IN
VACUO, AND THE PERCENTAGE BY WEIGHT REFERS TO THE FORMULA GIVEN.),
•
PerCent
By
Weight KOH
Density
NaOH NH.
Pel' Cent
By
Weight
Density
KOH \ NaOH NH.

II
1 1·0083 1·0095 0·9939 26 1·2489 1·2848 0·9040
2 1·0175 1·0207 0·9895 27 1·2592 - -
3 1·0267 1·0318 - 28 1·2695 1·3064 0·8980
4 1·0359 1·0428 0·9811 29 1·2800 - -
5 1·0452 1·0538 - 30 1·2905 1·3279 j)·8920.
6 1·0544 1·0648 0·9730 31 1·3010 - -
7 1·0637 1·0758 - 32 1·3117 1·3490 -
8 1·0730 1·0869 0·9651 ~3 1·3224 - -
9 1·0824 1·0979 - 34 1·3331 1·3696 -
10 1·0918 1·1089 0·9575 35 1·3440 - -
11
12
1·1013
1·1108
-
1·1309
-
0·9501
36
37
1·3549
1·3659
1·3900
-
-
-
13
14
1·1203
1-1299
-
1·1530
-
0·94,30
38
39
1·3769
1·3879
1·4101
-
-
-
15
16
1-1396
1·1493
-
1·1751
-
0·9362
40
41
1·399i
1·4103
1·4300 --
17 1·1590 - - 42 1·4215 1·4494 -
18 1·1688 1·1972 0'9295 43 1·4329 - -
19
20
21
22
1·1786
1-1884
1·1984
1·2083
-
1·2191
-
1·2411
-_
-
0·9229
0·9164
~
44
45
46
47
1·4443
1·4558
1·4673
1·4790
1·4685
-
1·4873
-
-
-
-
-
23 1·2184 - - 48 1'4907 H>065 -
24
25
1·2285
11.2387
1·2629
-
0·9101
-
49
60
1·5026
1·5143
-
1·5253
--
At 6. DATA ON THE STRENGTH OF AQUEOUS SOLUTIONS

OF THE COMMON ACIDS AND AMMONIUM: HYDROXIDE
Approximate Vol. required

to make 1
litre of
Reagent Per Cent. SpecifiC Nor- approx. N
by Weight Gravity mality solution.
MI.
Hydrochloric acid 35 1·18 11·3
Nitric acid 70-71 1·42 16·0
89
63 .
Sulphuric acid 96 1·84 36·0 28
Perchloric acid 70 1·66 11·6 86
Hydrofluoric acid 46 1-15 26·5 38
Phosphoric acid 85 - 1·69 14·7-44 23-68
Acetic acid 99·5 1,05 17·4 58
Ammonium hydroxide 27(NH.) 0·90 14·3 71
Appendix 791
A, 7. SAMPLES AND SOLUTIONS' FOR ELEMENTARY
'ANALYSIS
There are two main methods which may be used for the supply of
samples for analysis by elementary students, i.e., for students who are
just beginning the study of quantitative analysis. These are : -
1. Tbe'use of pure (e.g., A.R.) substances of known composition. This
method has been largely adopted in the present treatise, and in the
author's experience has given excellent results. It has been urged thiJ.t
the use of known amounts of material does not give the student practice
in discovering for himself the quantity of precipitant to be used and
solving similar simple problems which arise in. analysis. This is true,
but the writer feels that in the early ,stages of training it is far more
important that the student should first acquire the experimental
technique of quantitative analysis. The analysis of comparatively
simple " unknown" solutions and solids is better reserved for a later
stage when they will present compq_ratively little-difficulty. However,
many teachers prefer that quantitative exercises .should be performed
with solids or solutions of .. unknown" composition. These are dis-
cussed below. .
2: The use of standard solutions prepared from A.R. solids. A series
of solutions, each containing one element or radical to be determined,
is prepared of such concentration that. say, 20 to 3D ml. will provide
a convenient quantity for analysi~. For small classes the portion for
analysis is measured out with a graduated pipette; for large classes it is
cOl}venient to store the stock solution in a" bottle with a burette per-
manentlyattached (Fig. 2-25; the burette may also be of the automatic-
filling type as in Fig. 2-27). For each solution the instructor should
draw a large scale graph in which the volume of solution is plotted
against the quantity of the component present: this graph will provide
o a rapid means for checking the accuracy of the students' results.
Details of a few typical solutions are given below; others may be
similarly prepared (for list of A.R. reagents, see Section A, 9).
Aluminium: 70 g. of A.R. ammonium aluminium sulphate
(NH.J.SO.,AI.(SO.).,24H.O per litre.
Aplmonium: (i),15 g. of A.R. Jerrous ammonium sulphate
FeSO.,(NH.hSO.,6H.O per litre.
(ii) 4 g. of A.R. ammonium chloride NH.CI per litre.
Barium: 12 g. of A.R. barium.chloride BaCI.,2H.O per litre.
Bromide: 8 g. of A.R. potassium bromide KBr per litre.
Calcium: 20 g. of A.R. calcium carbonate CaCO. dissolved in a little
dilute hydrochloric acid and diluted to 1 litre.
Carbnnate: 30 g. of A.R. sodium carbonate Na.CO. per litre.
Chloride: 8 g. of A.R. potassium chloride KCI per litre.
Chf~mium: ~O g. of A.R. potassium dichromate K.Crp, per litre.
Cobalt: 8 g. of A.R. CXJbalt ammonium sulphate
CoSO.,(NH4).S0~,6H.O per litre.
Copper: 30 g. of A.R.,copper sulphate CuSO,,?H.O per litre.
Iodide: 7 g. of A.R. potassium iodide KI per litre.
192 Quantitative 11lorganic Analysis
Iron: (i) 50 g. of A.R. ferric ammoniull1asu1phate.
Fe.(SO.)3' (NH,hSO,,24HaO per litre,
(ii) 40 g. of A.R. ferrous ammonium sulphate
FeSO.,(NH.).SO.,6H.O per litre.
Lead: 12 g. of A.R. lead nitrate Pb(NOa) a per litre (a little nitric acid
should be present).
Magnesium: 25 g. of A.R. m3;gnesium sulphate MgSO.,7H.O per litre.
ManraneBe: 10 g. of A.R. potassium permanganate KMnO, per litre.
lIiIercury: 10 g. of A.R. mercuric chloride HgCI. per litre.
Niclte1: 7 g. of A.R. nickel ammonium sulphate
NiSO,,(NHo).SO,,6H.O per litre.
Phosphate: 1·2 g. of A.R. microc~smic salt Na(NH,)HPO,,4H.O p~-
litre (for the molybdate method).
Potassium: 7 g. of A.R. potassium chloride Kel per litre.
Silver: 14 g. of A.R. silver nitrate AgNO; per..J.itre:
Sulphate: 7 g. of A.R. potassium sulphate K.SO~ per litre.
Zinc: (i) 30 g. of A.R. zinc sulphate ZnSO,,7H.O per litre.
(ii) 40 g. of pure zinc ammonium sulphate
ZnSO,,(NH.hSO,,6H sO (see Sect!on IV, 34A) per litre.
A.8. ANALYSED SAMPLES
ANALYSED SAMPLES FOR STUDENTS IN
METALLURGiCAL ANALYSIS*
These are comparatively inexpensive analysed samples and are par-
ticularly useful for the training. of students in quantitadve analysis;
a sheet giving the proportions of the various components determined is
supplied. The use of many of these (" Ridsdale's Samples ") is sug-
gested in the present volume.
Sample No. Name. Constituents determined

0 Mild carbon steel Cd. C, P.
I Mild carbon steel Cd.C, Si, S, P, Mn .
2 Medium carbon steel Cd.C, Si, S, P. Mn .
3,21 High carbon steel Cd. C, Si, S, P, Mn .
22 Nickel steel Ni
3c Alloy steel Cd. C, Si, S, P, Mn, Cr, V, W .
4- Cast iron Cd. C, Si, S, P, Mn, Gr. C .
12b Foundry iron Cd. C, Si, S, P, Mn .
13b Hematite iron Cd.C, Si, S, P, Mn, Gr. C .
5b
- Brass Cu, Zn, Sn, Pb .
7 Phosphor bronze Cu, Zn, Sn, P, Pb .
19 Cupro-nickel Cu, Ni, Mn.
23 Copper (tough) As
10 Manganese bronze CU,Zn,Sn, Fe.
IOc Manganese bronze Cu, Zn, Sn, Fe, AI, Mn, Pb .
* Supplied by the Bureau of Analysed Samples, 3 Wilson Street, Middles-

brough, England. These 'were formerly called Ridsdale's Analysed Samples
for Students. and are referred to by this name throughout'the text.
Appendix 793
Analysed Samples for Students (n Metallurgical Analysis-cOIltilluetl
Sample So. Name Constituents determined

20 Aluminium alloy Cu, Zn, Ni, Mg, Fe, AI, Pb'.
32 Aluminium bronze Cu, Si, Fe, Al .
8b White metal Cu, Sn, Sb, Pb ,
lla Solder Sn, Pb.
9b Dolomite CaCO., MgCO" FesOs' AI.O"

Sia, .
I5b, I5c, 15d Limestone CaCO.,MgCO.,Fe.O.,AI.O., SiO •.
14b Fluor spar CaF., SiO. •
16a Coal (bituminous) Proximate analysis,S.
16b Coal (anthracite) Proximate analysis, S .
31 Thomas' phosphate
(basic slag) Complete analysis,
17 Iron ore (red) . Fe, SiO., P.

18 Manganese ore Mn, MllO,.
24 Portland cement Complete analysis.
25 Superphosphate W/S p.O" CaO .
26 Copper pyrites Cu, Fe, S.
27 Chrome iron ore C,;, Fe, Mg.
28 Iron pyrites Cu, Fe, S.
29b Felspar KIO, NasO, SiO•.
30 Zinc blende Zn,S.
41G Zinc concentrates Zn, S, Pb, Cu, Fe, As, Mn, Cd,
CaF•.
42G Lead concentrates Pb, Zn, Ag, Cu, Bi, Au, Sb: As,
Cu, Fe, Mn, S .
43G Tin concentrates Sn, W, As, Cu, Pb, Bi, S, Fe, Mn.
44G Iron pyrites Cu, Se, As; Co, Zn .
45G Burnt pyrites Cu, Fe, S.
46G Cu.Pb matte Cu, Pb,'Ag.
47G Cu·Ni speiss Cu, Pb, Ni, Co, As, Ag .
BRITISH CHEMICAL STANDARDS*

Description SPecial features Elements determined
Carbon steels
U AI II
Fe standard for C,Si,S,P,Mn,As,Ni,Cu,Fe.
K.Cr.07
"e" C standard only C.
tiD" Free cutting, high S C,Si,S,P,Mn.
U'El II
C and Mn standard C,Mn.
(IN" Complete analysis C,Si,S,P,Mn,As,Ni,Cu.
"P" High S and P stan· C,S,P,Mn.
dard
"M'· C and Si standard ~,Si.
u 0 1 ., C,Si,S,P,Mn,As,Ni,Cr.
Complete analysis
"HI" C, S, and P stan· C, 5, P.
dard
"Ft. Complete analysis C,Si, S,P, Mn .
U j" Complete analysis C, Si, S, P, Mn.
• Supplied by the Bureau of Analysed Sam'ples Ltd., 3 Wilson Street,

Middlesbrough, England.
British Chemical Standards-continued
,.
Description Special features Elements determined
1----------------1----------------1-------------------------
Carbon ateela-cont.
fiR" C, Sand Mn stan- C, S, Mn~
dard
If 51" C and P standard C, P.
" U" C,Mn and Ni stan- C, Mn, Ni.
dard -
'Bickel 8teel
If. '["t" Ni, C and Mn stan- C, Mn, Ni. \
Allo18teels dard
V"
II Cr-V steel C,Si,S,P,Mn,Cr,V .
"W'· Cr-V-W-Co steel C,Si,S,P,Mn,Ni,Cr,V,Co,W.
"~W2" Cr-V-W-Co·Mo C,Si,S,P,Mn,Mo,Ni,Cr,V,Co,W.
steel
"E" Cr-Ni-Mo steel C,Si,S,P,Mn,Mo,Ni,Cr .
Stainless steel Cr, Ni
CaatIroDa
Hematite" AS" -Complete ana1ysis Si,S,P,Mn,Ti; Total C, Gr: C .
Hematite" 206 " High silicon Si,S,Mn, Total C, Gr. C .
Hematite" B " Low phosphorus, S,P,,t\s,Ti, Total C, Gr. C.
titaniferous
Foundry .. D' " SI, P and Mn stan- Si,P,Mn.
dards
High duty" G" Complete. analysis, Si, S,P,Mn,As,Ni,Ti, Total C,
high S Gr.C.
Alloy" K" Complete analysis, Si,S,P,Mn,Mo,Ni,Cr, Total C,
Ni, Cr, Mo Gr. C.
Austenitic .. L " High Ni-Cu-Cr Si,S,P,Mn,~i,Cr, Total C, Cu.
Perro Alloys
Ferro-W, No. 203 Tungsten, low C W.
Ferro-Mo. No. 200 Molybdenum, low C Mo .
Ferro-Cr. No. 203 Chromium, low C Cr, C, S.
andS
Ferror-Cr. No. ~04 Chromium, high C Cr, C, S.
andS
Ferro-Mn Manganese • Mn .
Ferro-V. No. 205 Vanadium V.
Ferro-Ti. No. 201 Titanium, low C Ti.
Bl).sic slag.".A " P,O, and partial P,O.,SiO.,CaO,MgO, Fe
analysis
Om
Iron ore" A " Complete analysis Fe,P,CaO,MgO,Alkalis,Al.O.,
and K,Cr.0 7 TiO.,SiO.,S, Cd. H.O, CO •.
standard
Mang. Ore" A " Mn,O, SiO.,P, Fe Fe,P,Mn,O,SiO •.
Non-ferrous AnOYl
WRite metal .. A " Complete analysis Pb,Sb,Sn,Cu,Fe,Bi,As,Zn
White metal" B " Complete analysis Pb,Sb,Sn,Cu,Fe,As,Zn .
Mang. brass .. E " Complete analysis Pb,Sn,Cu,Fe,Mn,Zn,Ni,Al .
Cupro-nickel .. A " Complete analysis Si,Cu,Fe,Mn,As,Zn,Ni,Mg,Al.
Alum. alloy" A .. Complete analysis Pb,Sn,Cu,Fe,Zn,Ni,Mg,Al
Alum. alloy" B " Complete analysis, Si. Fe, Al
highS
Bronze" A" Complete analysis Pb,Sb,Sn,Cu,Fe,As,Zn,P .
Bronze" C" Complete analysis Pb,Sb,Sn,Cu,Fe,As,Zn,Ni,P, S
Appendix 795
British Chemical Standards---cO'ntinued
Benzoic acid Purity 99'9%
Sodium oxalate Purity 99'97%
Tin Purity 99'9%, m.p. 231·9·C
Lead Purity 99'98%, m.p. 327 '3°C
Zinc Purity 99'94%, m.p. 419'4°C
Aluminium Purity 99'82%, m.p. 659·2·C
Sodium' chloride' Purity 99-96%, m.p. SOI°C
Copper I Purity 99'91 %, m.p. 1083°C
BRITISH STANDARDISED STEEL SAMPLES·

The following s~dards are at present available :-
No. I.-Sulphur (S = 0·027%).
No. 2.-Sulphur (S = 0'011%).
No. 3.-Phosphorus (P -= 0'029%).
No. 5.---Carbon (C = 0-65%). Acid O.-H. steel.
No. 6.-Carbon (C = 0-10%). Basic O.-H. steel.
No. 8.-Carbon (C = 0·27%). Acid O.-H. steel.
No. 9.-Carbon (C = 1'09%). Acid O.-H. steel. •
No. H.-Manganese (Mn = 1'09%). Acid O.-H. steel.
No. 12.-Cast iron standard sample (Si = 2·22%, P = 1·14%, Mn = 0·50%
S = 0'075%).
STANDARD SAMPLES ISSUED BY THE U.S.

NATiONAL BUREAU OF STANOARDSt
I
Sample
No.
Name
I Constituents detel'mined
Steels
8e Bessemer,O'IC C,Mn,P,S.Si (Cu,Ni,Cr.V) .

9c Bessemer, 0·2C C,Mn,P,S,Si (Cu,Ni.Cr,V.Sn).
lOd Bessemer, O'4C C.Mn,P,S,Si (Cu,Ni,Cr,V,N) .
22b Bessemer, 0·6C C,Mn.P,S.Si (Cu,Ni,Cr,V,Sn).
15b B.O.H.,O·IC C,Mn,P,S,Si (Cu,Ni,Cr,V,As).
lld B.O.H., O'2C C,Mn,P,S,Si (Cu,Ni,Cr,V,As).
12d B.O.H., O'4C C,Mn,P,S,Si (Cu,Ni,Cr,V,As).
13c B.O.H.,O·6C C,Mn,P,S,Si (Cu,Ni,Cr,Al,Sn).
14<: B.O.H., O·se C,Mn,P,S,Si (Cu,Ni,Cr,V,Al,Al.O.).
16c B.O.H., l'OC C,Mn,P.S,Si (Cu,Ni,Cr,V).
19c A.O.H., 0·2C C,Mn,P,S,Si (Gu,Ni,Cr,V).
20c A.O.H., 0'4C C,Mn,P,S,Si (Cu,Ni,Cr,V) .
21c A.O.H.,O·6C C,Mn,P,S,Si (Cu,Ni,Cr,V).
34a A.O.H.,O·se C,Mn,P,S,Si (Cu,Cr,Mo).
35a A.O.H., l'OC C,Mn,P,S.Si (Cu,Cr,Al).
• These are issued jointly by the Iron and Steel Institute and the National
Physical Laboratory, alld are obtainable from the National Physical Labora-
tory, Metallurgy Department, Teddington, Middlesex, England.
t Obtainable from the U.S. Department of Commerce, Bureau of Standards,
Washington. D.C.• U.S.A.
Sa",pl~
Name Constituents determined
No.
I I
Stee~ontinued
51a Electric furnace, 1·2C C,Mn,P,S,Si (Cu,Ni,Cr, V,Sn).

65a Acid electric C,Mn,P,S,Si (Cu,Ni,Cr,V,Sn) .
100 Medium manganese C,Mn,P,S,Si (Cu,Ni,Cr,V) .
105 High sulphur C
30c
-
Chrome-vanadium C,Mn,P,S,Si;Cr,V (Cu,Ni) .
32b Chrome-nickel C,Mn,P,S,Si,Cr,:t{i. (Cu) _
33b Nickel C,Mn,P,S,Si,Ni (Cu,Cr,V) .
50a Cr-W-V C,Mn,P.S,Si,W,Cr,V (Cu.Mo,Sn) .
-
72a Chrome-molybdenum C,Mn,P.S.Si,Cr,Mo (Cu, V) .
106 Cr-Mo-Al (Nitralloy "G") C.Mn,P,S.Si.Cr,Mo,AI (Cu,Ni,V,As.
N).
III Ni-Mo-Cr (SAE 4615) C,Mn,P.S,Si,Cr,Ni.Mo (Cu,V,As) .
73
lOla
. Stainless
18Cr.8Ni .
C.Mn.P.S,Si,Cr (Cu, V.Mo) .
C,Mn,P,S.Si,Cr,Ni (Cu,V.N,Mo) .
121 ISCr.9Ni (Ti bearing) C.Mn,P.S.Si.Cr,Ni,Ti (Cu.V,N,Mo) .
123 lSCr,IlNi (Cb bearing) C.Mn,P,S,Si,Cr,Ni,Cb (Cu,V) .
125 High silicon C,Mn,P,S,Si,Cu,Ni,Cr,Al (V.Mo.Ti,
Sn).
Irona -
4e Cast iron C,Mn.P,S,Si,Ti (Cu,Ni,Cr.V) .
5g Cast iron C,Mn.P,S.Si,Ti (Cu,Ni.Cr,V) .
-6d Cast iron C,Mn,P,S.Si,Ti (Cu,Ni,Cr,V,Sn) .
7c Cast iron C,Mn,P.S.Si,Ti (Cu,Ni,Cr,V).
55a Ingot iron C,Mn,P,S.Si,Cu (Ni,Cr,N,AI.AI1O.,
Co.Sn) .
82 Cast iron C;~'tn,P.S.Si,Cr,Ni (Ti,Cu,V) .
107 Ni-Cr cast iron C,Mn.P,S,Si,Ni,Mo.Cr (V.Cu.Ti) .
115 Ni-Mo cast iron C,Mn,P,S,Si,Ni,Cr,Cu (V,Mo) _
122 Cast iron (car wheel) C.Mn.P,S,Si,Ni.Cr,Cu (V. Mo).
Steel-making aDoJS
57 Refined silicon Complete analysis.

58 Ferro-silicon (75% Si) do.
59 Ferro-silicon (50% Silo do.
61 Ferro-vanadium (high C) do.
66 Spiegeleisen do.
67 Manganese steel do.
68 Ferro-manganese do.
75 Ferro-tungsten do.
90 Ferro-phosphorus P
71 Calcium molybdate Mo. Fe, Ti
ll6 Ferro-titanium (low C) Ti. C, Si, V, Cr. AI.
117 Ferro-titanium (high C) Ti, C. Si. V, Cr, Al .
Appendix 797
Sample
No.
I Name I CnnstifflBnts determined
~------------------~----~---------------------
Non-ferrous alloys
86 AI-base casting alloy Complete analysis.
53a Bearing metal, Pb-base do.
54a Bearing metal, Sn-base do.
63 Bearing metal, phosphor
bronze do.
37b Brass, sheet do.
52 Bronze, cast do.
124 Ounce metal do.
62 Bronze, manganese do.
94 Zn-base, die casting alloy do.
95 Do. do.
96 Do. do.
Ores
69 Bauxite Complete analysis.
26 Iron are, Crescent Al.O., Cao, MgO .
29 Iron ore, Magnetite Complete analysis.
28 Iron are, Norrie Mn(low).
27b Iron are, Sibley SiO" P, Fe.
25b Manganese ore -Mn, available oxygen.
56a Phosphate rock
(Tennessee) P.O.,Fe.O.,Al.O., etc.
120 Phosphate rock (Florida) P.O., Fe.O" AI.O a, etc.
2a Zinc are Zn
113 Zinc orc (Tri-State Con-
centrate) Zn
Ceramic materials
104 •Burned magnesite Complete analysis
76 Burned refractory
(40% Al.O.) do.
(60% Alp.) do.
(70% AI,O.) do.
103 Chrome refractory CrIO •• SiO•• AlsO •• FeO.CaO.MgO
97 Clay. flint Complete analysis.
98 Clay. plastic do.
70 Feldspar (potash) do.
99 Feldspar (soda) do.
79 Fluorspar I do.
la Limestone, argillaceous do.
88 Limestone. dolomitic do.
92 Glass. low boron B,O,.
93 Glass. high boron Complete analysis.
89 Glass. lead-barium do.
91 Glass, opal do.
80 Glass. soda-lime do.
81 Glass sand Fe,O"AIIO •• TiO•• ZrO•• CaO,MgO
102 Silica brick Complete analysis.
112 Silicon carbide do.
798 Quantitative Inorganic Analysis,
Sample I Name Constituents determined

No.
I
Fineness standards .
46r Cement (normal)
m c
.
Cement (extra fine)
Cement (turbidimetric standard)
Melting-point standards
44c Aluminium 660·15°C.

45a Copper 1,083°C.
49a Lead 327·35°C.
42c Tin 231·87°C.
43d Zinc 41IH2°C.
Chemidala
84 Acid potassium phthalate Acidimetric value.

(99'97%)
3ge Benzoic acid (99'99%) Acidimetric and calorimetric value.
40c Sodium oxalate Oxidimetric value.
83 Arsenk trioxide (99·97%) Oxidimetric value.
3Sb Naphthalene Calorimetric value.
17 Sucrose (cane sugar) Calorimetric and saccharimetric
value.
41 Dextrose (glucose) Reducirtg value.
Thermo-elecmc standards "-

-
118 Alumel wire, no. 8 gage E.m.f .. N.B.S. Pt no. 27, (10 to
1300°C.
119 Chromel wire, no. 8 gage E.m.f. N.B.S. Pt. no. 27, O· to
1300°C.
A,9. ANALYTICAL REAGENTS

In the following tables a list of the various analytical reagents at
present available from British and American sources is given. The
products of the following firms are listed :-
1= Hopkin and Williams Ltd., 16St. Cross St., London, E.C.2, England.
2 = Mallinckrodt Chemical Works, 72-74 Gold Street, New York,
N.Y., U.S.A.
3 = Baker and Adamson, Division of the General ChemiCal Co., 40
Rector Street, New York, N.Y., U.S.A.
4 = J. T. Baker Chemical Co., Phillipsburg, N.J., U.S.A.
S == The Coleman and Bell Co., Norwood, Ohio, U.S.A.-
6 = Merck and Co., Rahway, N.]., U.S.A.
-This firm terms most of its products of analytical reagent quality
.. Chemically pure" (C.P.); the~ have been incorporated in the tables. •
Appendix 799
In view of the limited number of analytical reagents ·at present
available of purely'British origin, 'Messrs. Hopkin and William'! Ltd., of •
St. Cross Street, London, E.C.2, England, have, at the author's request,
made arrangements to supply any of the Atperican products at short
. notice.
INORGANIC COMPOl;lNDSt
Aluminium
Metal Al 3 Fluoride AIF a,xH.O 3,4
Acetate, basic
Ammoniul!l sulphate
AI(SO')3,(NH,).SO,.
.
3,4*, ti Nitrate AI(NO.)., 9H.O 2.3,4,5,6
Oxide, ignited AlIO,
Potassium sulphate
~,3,4. 6
24H.O
Chloride AlCl.6H.O
- 1. 2. 3. 4. 5, 6
2,3,4,5,6
Al.(SO,).,K.SO,,24H.O 1,2,3,4.'5,6
Sodium sulphate
Chloride, anhydrous A).(SOJ.,Na.SO,,24H.O 3,4,5
AlCI. 2,3,4,6 Sulphate AI.(SO,).,18H.O 1,2,3,4,5,6
Ammonium ,
Hydroxide s.g. 0·88-0·90 I, 2, 3. 4, 5. 6 Metavanadate NH, va. 1,5 ...
Acetate NH,C.H.O. 1.2,3,4,5,6 Molybdate
Arsenate 4 (NH,).Mo,O.,,4H.O 1,2,3, 4. 5, 6
Bicarbonate NH,HCO. 1,3,4,5 Nitrate NH,NO. 1,2,3,4,5,6
Bichromate (NH,).Cr.07 2,3,6 Oxalate (NH,).C.O,.H~O 1.2,3,4,5,6
Bisulphate NH.HSO. 2,3,1>,5,6 Perchlorate NH.ClO. 4,5
Bitartrate NH,.HC,H,O. 4,5 Persulphate. (NH,) ,S.O. 1. 2, 3, 4, 5, 6
Borate 4 Phosphate, primary.
Bromide NH,Br 2,3,4,5,6 NH,H.PO, 2.3,4,5,6
.. Carbonate .. I, 2, 3, 4, 5, 6 Phosphate, secondary,
Chloride NH,Cl I, 2, 3, 4, 5, 6 (NH,).HPO, I, 2. 3, 4, 5, 6
Chromate (NH,)zCrO, 2,3,4,5,6 Sulphate (NH,).SO, 1,2,3,4, Ii. 6
Chromium sulphate Sulphide solution, light
(NH.).SO"Cr.(SO,) .' (NH,).S + Aq. "2,3,4.5
24H.0 3 Sulphide solution, yellow
Citrate (NH.).H.C.H,o, 2,3,4,5,6 (NH.).S. + Aq. 1,2.3,4,5.6
Fluoride NH,F 1,3.4 Tartrate (NH,) •. C,H,O. 3,4,5.6
Formate HCOONH, 3,4,6 Thiocyanate NH,CNS I, 2, 3, 4, Ii, 6
IodideNH.I 2,3,4,5
Antimony
Metal Sb 2. 3, 4 Sulphate Sb.(SO,). IS
Pentoxide Sb.O. 5 Tribromide SbBr. 5
Pentachloride SbCI. 3,4,5 Trichloride SbCI. 1,2,3,4,5,6
Potassium tartrate Trioxide Sb.O. 3.4,5
2K(SbO)C,HP.;H.0 1*,3,4,5
t All organic compounds con'taining inorganic cations will be included under

inorganic compounds.
• Throughout these tables an asterisk indicates that the anhydrous compound is
also available.
Anenic.
Arsenic As
Arsenic acid HaAsO,
5
2,3 .. 6
I Trichloride AsCI,
Trioxide AsIO.
:I
1,2,3,4,5,6
Pentoxiae AsIO, 2,3,4,5
Barium
Acetate Ba(C,H,Oz)z,H,O 1,2,3,4,5,6 Nitrate Ba(NO')1 I, 2, 3, 4, 5, 6
Bromate Ba(BrO,h,H.O 2 Oxalate BaC.O"H,o ",5
Bromide BaBr., 2H.O 2,5 OxideBaO 3
Carbonate BaCO. 1,2,3,4,5,6 Phosphate Ba.(PO,). 4,5
Chlorate Ba(CIO.J.,H.O 3,4,5,6 Sulphate BaSO, 2,3,4,5,6
Chloride BaCI" 2H.O I, 2, 3, 4, 5, 6 Sulphite BaSO. 4
Chromate BaCrO, 4,5 Tartrate Ba.C,H,O. (j
Dioxide, anhydrous BaO, 2,3,4,5,6 Thiocyanate
DioXide BaOI,SH.O 4 Ba(CNS).,3H.O 3
Fluoride BaF. 3 Thiosulphate BaS.o.,
Hydroxide Ba(OH)., H,O 4
SH,O 1, 2, 3, 4, 5, 6
Bismuth
MetalBi 2.3,4,5,6 Sub-nitrate BiONOa,HIO 2, 3, 4, 5
Nitrate Bi(NO.) ,,5H.O 2,3,4,5,6 Trichloride BiCI. 2,3,4,5,6
Oxychloride BiOCI 3,4 Trioxide Bi.a.. 3,,4
Sub-carbonate
Bi.O.,CO"H,O 4,5,6
Boric Acid H.BO, 1,.2,3, .f, 5 ,
Boric anhydride 6• .
Bromine Br. 1,2, 3,4,5
Oadmium
Acetate cJ(C.H,Oal., IodideCdI. I, 2, 3, 4, Ii, 6
2H.O " 2,3,4,5,6 Nitrate Cd(NO.l.,4H.O 2,3,4,5,6
Bromide CdBr.,4H.0 3,4,5,6 OxideCdO 3,4,5
Carbonate CdCO. 2, 3,~, 5,6, Potassium iodide
Chloride 2CdCI,,5H,0 2,3,4,5*,6 CdI.,2KI,2H.O 2
Hydroxide Cd(OH). 3,4,5 Sulphate 3CdSO,,8H.0 1,2,3,4*,5, II
calcium
Acetate Ca(C,H.Oa)"H,O 2, 3, 4, 5, 6 Nitrate Ca(NO.l l ,4H.0 2,3, 4,Ii, 6
Arsenate 4 Ox8.late CaC,o"HaO 3,4,5
Bromide CaBrI 3,4,5 OxideCaO 3,6
Carbonate CaCO. 1,2,3,4,5,6 Phosphate, primary,
Chloride CaCI,,6H,0 1,2,3,4,5,6 Ca(H,PO,l.,H.O • 2,3,4,5; 6
Chloride, anhydrous Phosphate, secondary
(96%lICaCl, 2,3,4,5,6 CaHPO,,2H,O 2,3,4,5,6
Chromate CaCrO,,2H.0 4,5 Phosphate, tertiary
Citrate Ca.(C.H,o,l. Ca,(PO,l: 2,3,4,5,6
4H.0 3,4,5 Sulphate CaSO,,2H.O 2,3,4,5,6
Fluoride CaF, '3,4;5 Sulphide CaS 3
Formate Ca(HCO,). 4,5,6 Sulphite CaSO.,2H.0 3,4,Ii
Hydroxide Ca(OHls 2,3 'Tartrate Cac,H,0,,4H.0 3,4,5
Iodide Cal, 4 Tungstate 5
'Cerium nitrate S
Ohloropiatinic acid H,PtCl.,6H.o 2,3
Appendix 801
Chromium
Acetate Hydroxide CrsO.,xHsO 4
Cr(CzHaO.J "xH,O 2,3,4,5 Nitrate Cr(NO a).,9H.O 2,3,4,5
Ammonium sulphate Potassium sulphate
Cr,(SO,)., (NH,).SO" Crz(SO,).,K,SO,,24H,O I, 2, 3, 4, 5, 6
UH.O 4 Sesquioxide Cr.O. 3,4,5
Bromide CrBr.,xHp 3, 5 Sulphate Cr,(SO,).,xHsO 2,3,4,5,6
Carbonate, basic 5 Trioxide CrD, 1,2,3,4, 5, 6
Chloride CrCI.,6HP 2,3,4,5
Cobalt
Metal (118-99%) 3,4,5 Chloride CoCI.,6H zO I, 2, 3, 4, 5, 6
Acetate CO(C,H.O,)" Chromate CoCrO, 5
4Hp 2,3,4,5,6 Nitrate Co(NO.),,6H,O I, 2, 3, 4, 6, 6
Ammonium sulphate Oxalate CoC.0.,2HP 4,6
CoSO" (NH,).SO" BH.O 4,5 Oxide, black CosO. 3,4
Carbonate COCO, 3,4,5,6 Sulphate CoSO,,7HzO 1,2,3, 4, 5, 6
Copper
MetalCu 1,2,B Ferrocyanide CuIFe(CN). 6
Acetate Cu(C,H.O,).,H,O 1,2,3,4,5,6 Nitrate Cu(NO.)s,3H.O 1,2,3,4,5, 6
Aceto-arsenite Oxalate 2euC,0"HI O 4,6
(CuO.As,O.).(CsH.O.)s Oxide (ie) CuO I, 2, 3, 4, 5, 6
Cu 4 Oxide (ous) CulO 3,4,5
Ammonium chloride Phosphate Cu.(PO,)s,
CuClz,2NH,CI,2H zO I, 2,3, 4, 5, 6 2HaO 4
Ammonium sulphate Potassium chloride
CuSO,,(NHJ zSO,,6HP 3,4,5 CuCla,2KCI,2HsO 3,4,6
Borate (ic) 5 Potassium sulphate
Bromide (ie) CuBr, 2,3,4,5,6 CuSO"KaSO,,6H,0 3,4
Bromide (ous) CUsBr, =!,4,5 Sulphate CuSO,,5HsO 1. 2, 3, 4, 6, 6
Carbonate, basic Sulphate, anhydrous
ca. CuCO.,Cu(OH). 2,3,4,5,6 CuSO, 2,3,4,6,6
Chloride (ic) CuCI,,2H.o 1,2,3,4*,5,6 Sulphide (ic) CuS 2,3,4,5
Chloride (ous) Cu.Cl. I, 2, 3, 4,-5, 6 Sulphide (ous).CusS 2,6
Cyanide (ous) Cus(CN)z 3,4 Tartrate CuC,H,0,,3HP 4
Devarda's metal 6
Eshka's mixture 2MgO : 1Na,CO. 1, 2, 3, 4, 5, 6
Iron
Ferric compounds Ferrous compounds
Ammonium oxalate Ammonium chloride
Fe(NH,) • (C,O,) .,3H.0 4,6 FeCI•• 2NH,CI,H.0 4,5
Ammonium sulphate Ammonium sulphate
Fe.(SO,). (NH,)zSO" FeSO,,(NH,hSO,,6H.0 I, 2, 3, 4, 5, 6
24H,O I, 2, 3, 4. 5, 6 Chloride FeCI s, 4H,0 2,3;4,6,6
Chloride FeCI.,6H.0 1, 2, 3, 4, 5, 6 Oxalate FeC.0.,2H.0 5
Chloride'; anhydrous Phosphate 3.5
FeCI. . 1,3,5 Sulphate FeSO" 7HaO 1. 2, 3, 4, IS, 6
Nitrate Fe(NO.)3,9HzO 2,3,4,5,6
Oxide FesO. 3,4
Phosphate FePO,,4H.O 3,4,5
Sulphate Fe.(SO,) .,xH.O 2,3,4.5,6
DD
802 Quantitative I norgll1tic A 1talysis
Fuaion mixture 1
Gold Chloride HAuCI,,3H I O 2.3,6
HydraziDe sulphate N.H"H I SO, 1,5, fl
Hydrogen peroxide, 3% H 2 O, 1,2,:l,4
Hydrogen peroxide, 6% HP. 1
Hydrogen peroxide, 30% H 2O. 4
Hydrobromic acid 46-48% HBr 1,2,3,4,5,6
Hydrochloric acid 35--37% HCl 1, 2, 3, 4, 5, 6
HydroDuorie acid 40-48% HF I, 2, 3, 4, 5, 6
Hydriodicacid47% HI 2,Il
Hydriodie acid 54-56% and 64-68% HI 1,6
HydroDuosilieie acid 27-30% H.SiF. :J, 4, 5, fl
Hydroxylamine hydrochloride NH.OH,HCl 1,3,4,5,6
Hydroxylamine sulphate (NH,OH)"H.SO. 3,5
Iron wire 99'9% Fe 2,3,6
Iodine I. 1,2,3.4,5,6
Iodie acid HIO 3 I, 2, 3, 4, 5, 6
Iodine pentoxide 1,0. 2. 6
Lead
Metal Pb 1,2,0 Chromate PbCrO, 2.3,4,6,6
Acetate Pb(C.H.O,).3H.D I, 2. 3. 4. 5, 6 Dioxide PbO, 1. 2, 3, 4. 5, 6·
Acetate, basic Nitrate Pb(NO.). 1,2,3,4,5
ca.5Pb(C.H.O,).,2PbO 1,2,3,5,6 Oxide, yellow, PbO I, 2. 3, 4, 5, 6
Arsenate PbHAsO, 4 Oxide, red, Pb.D, 2,3.4.5,6
Borate Pb(BO.).,H.O 4 Oxalate PbC,O, 4
Carbonate, basic Phosphate Pb.(PO.l, 2,4,5
ca. 2PbC0 3 ,Pb(OH). 2,3,4,5,6 Sulphate PbSO. 2,3,4,5,6
Chloride PbCl. 2,3.4,5,6 I Tartrate PbC.H.O, 4
Lithium
Acetate 5 Fluoride LiF 3,4
Bromide LiBr 5 Nitrate LiNO. 2,3,4,5,6
Carbonate Li.CO. 2,3,4,5 Oxalate 5
Chloride LiCI 2,3,4,5,6 Sulphate Li,SO.,H.O 2,3,4,5,6
. Magnesium
Metal, turnings Mg 2 Citrate .4
Acetate Fluoride MgF, 4
Mg(C.H.0.),,4H,O 2,3,4,5,6 Nitrate Mg(ND.l,,6H.0 2,3.4,5,6
Ammonium chloride OxideMgO 1, 2. 3, 4, 5, 6
MgCl"NH,CI,6H,0 1,2,3,4,5 Oxalate MgC,O,,2H.0 4
Bromide MgBr.,6H,O 3,4,5,6 P.erchlorate Mg(CIO.). 6
Carbonate, basic Phosphate, secondary
ca.3MgCO.,Mg(OH)., MgHPO.,3H,O 4
3H.0 2,3,4,5,6 Phosphate. tertiary
Chloride MgCI.,6H.0 1, 2, 3, 4, 5, 6 Mg.(PO,jz,8H.0 2,4
Chromate 4 Sulphate MgSO., 7H,O 1,2" ,3,4,5",6
Manganese
AcetateMn(C.H.0.).4H,0 4, 5 Dioxide MuO, 3,6,6
Bor~te 4, Nitrate Mn(NO.).
Bromide 5 50% solution 2,3,4,5,6
Carbonate MnCO. 3,4,5,6 Phosphate 4
Chloride MnCI.,4H.0 1, 2, 3, 4, 6, 6 Sulphate MnSO.,4H.0 1,2.3,'4,6,6
Appendix 803
Jderclll'J
MetalHg 1, 2, 3,4, 5, 6
"Mercuric compounds I Mercurous compounds
Acetate Hg(C.H.O,). 2,3,4,5,6 Chloride Hg.CI. I, 2, 3, 4, 5, 6
Bromide HgRr. 2,3,.4,0,6 Iodide HgaI. 5
Chloride HgCl. I, 2, 3, 4, 5, 6 Nitrate Hg.(NO.).,2H.O I, 2, 3, 4, 5, 6
Cyanide Hg(CN). 2,4,6 Sulphate Hg.SO, 2,3,4,6
Iodide HgI. 2,3,4,5,6
Nitrate Hg(NOs).,xH.O 2,3,4,5,6
Oxide, red HgO I, 2, 3, 4, 5, 6
Oxide, yellow HgO 2,3,4,5,6
Sulphate HgSO, 2,3,4,5,6
Sulphide HgS 4
Molybdenum trioside MoO. 1, 2, 3, 4, 6, 6

Molybdic acid (86% MoO.) 1,2,3,4,6, (I
Hickel
MetalNi 2,4 Chloride NiCI.,6H.O I, 2, 3, 4, 5, 6

Acetate Ni(C.HsO.)., Nitrate Ni(NO.l.,6H.O 2,3,4,5,6
4HaC 3,4,5,6 Oxalate NiC.O.,2H.O 4,6
Ammonium sulphate Oxide, grey-black (ic) 4,6
NiSO,,(NH,),SO,,6HP 2,3,4,0,6 Oxide, green (ous) NiO 4
Bromide 5 Phosphate Ni.(POJ.,
Carbonate, basic 7H.O 4,6
ca. 2NiCO.,3Ni(OH)., Sulphate NiSO,,6H.O 1, 2, 3, 4, 5, 6
4HaC 3,4,5,6 Tartrate 4
!litric acid, sp. g(. 1'42, 69-72% HNO. I, 2, 3, 4, 5, 6

Nitric acid, sp. gr. 1'5, fuming 89-95% HNO. 1,3,4,5,6
Oxalic acid, cryst., H.C,O.,2H.O 1, 2, 3, 4, 5, 6
Oxalio acid, anhydrous H.C.O, 3,4,5
Palladium chloride 5
Perchloric acid, 20% or 60% HCIO, 1, 2, 3, 4, 5, 6
Perchloric acid, 70-72% HCIO,
Phosphoric anhydride p,O.
Phosphoric acid, ortho, 85% HaPO,
. 2,3,4,5,6
2, 3, ~ 5, 6
I, 2, 3, 4, 5, 6
Phosphoric acid, meta, sticks 2,3,5,6
Phosphorous acid, cryst. H.PO. 3
Phosphorus oxychloride POCI. 2,3,5,6
Phosphorus pentachloride PCl, 2.3,5,6
Phosphorus trichloride PCl. 2,3,4,5,6
Phosphomolybdio acid P .o.,20MoO.,xH.O I, 2, 3, 4, 5, 6 "-
Phosphotungstic acid P aC,,24MoO.,xH.O 1,2, 3, 4, 5, 6
Potassium
Acetate KC.H.O. 2,3,4,5,6 Bichromate K.Cr.o, 1,2,3,4, 5, 6

Ammonium sulphate Binoxalate KHC.O, 3,4,6
K.SO" (NH,).SO. 4,5 I, 2, 3, 4, 5, 6
Bisulphate KHSO.
Areenate, primary
KH.AsO, 3,4,5,6 Bitartrate KHC,H,O. 1,2,3,4,5, 6
Arsenite KAsO I 3,4,5,6 Bromate KBrO. 1,2,3,4, 5, 6
Bicarbonate 1, 2, 3, 4, 5, 6 Bromide 1, 2, 3, 4, 5, Ii
Potassium-continued
Carbonate, anhydrous Nitrite KNO, 2,3,4,5
KsCO, I, 2, 3, 4, 5, 6 Oxalate K,C,O"H,O 1,2,3,4,5,6
Carbonate, cryst., Perchlorate KClO, 3,4,5.6
2K,CO a,3HIIO 4,5 Periodate KIO, 3.4
Chlorate KClO. 1.2.3.4.5.6 Permanganate KMnO, I, 2, 3, 4, 5, 6
Chloride KCl 1,2,3,4,5 Persulphate K,S,OS 1, 2, 3, 4, 5, 6
Chloroplatinate K.PtCl, 5 PQosphate, primary
Chromate K.CrO. 1, 2, 3, 4, 5, 6 KHsPO, 1, 2, 3, 4, 5, 6
qtrate K SC8H,07,H,O 1,2, 3, 4, 5, 6 Phosphate, secobdary
Cyanate KCNO 3,4,5 K,HPO, 2,3,4,5,6
CyanideKCN 1, 2, 3, 4, 5, 6 Phosphate, tertiary
Ferricyanide'K,Fe(CN), 1, 2, 3, 4, 5, 6 K.PO, 2,3,4,5,6
Ferrocyanide Pyroantimonate 4,5
K,Fe(CN).,3HP 1,2,3,4,5,6 Pyrophosphate
FluorideKF 2,3,4,5,6 K,P,07,3H,O 4,5
Formate HCOOK 4,5 Sodium tartrate
Hydrogen phthalate ' KNaC,H,Oe,4HP 1, 2, 3, 4, 5, 6
KHCaH,O. 1,2,;1,4,5,6 Sulphate K.SO, I, 2, 3, 4, 5, 6
Hydroxide KOH I, 2, 3, 4, 5, 6 Sulphite KaSO. 3,4,5
Iodate KIO. 1,2,3,4,5,6 Tartrate 2K~C,H,08,H,O 2,3,4,6,6
Iodide KI 1, 2, 3, 4, 5, 6 Tetraborate
Mercuric iodide 4 K t B,O,,5H,O 3,4
Meta~isulphite 1,2, 3,4,5,6 Tetroxalate
Molybdate 5 KHC.O"H aCsO.,2H.O 1, 4, 5, 6
Nitrate KNO. 1,2, 3,4,5,6 Thiocyanate KCNS I, 2, 3. 4, 5, 6
Silicic acid. powder 2,3.4,5.6

.
Silver
Metal. powder Ag 2 Lactate AgC.H.O.,H.O 3

Bromid:l AgBr 5 Nitrate AgNO s 1,2.3,4,5,6
Carbonate Ag.CO. 2,3 Oxide Ag.O 3
Chloride Agel 2,3,4,5 Sulphate Ag,SO, 2,3,4,5,6
Iodate 6
Sodium
Metal, Na 2,3,4,5,6 Bitartrate
Acetate NaC,H.O •• 3H.O I, 2, 3,4,5,6 NaHC,Hp.,H,O 2,3,4,5,6
Acetate, anhydrous Borate NaB,07,IOH 2O I, 2,3,4,6, 6
NaC,H.O. I, 2, 3, 4, 5, 6 Br.omate NaBIO. 2,3,4,5,6
Aluminate NasAI.O, 3 Bromide N aBr 2,3,'4,5,6
Ammonium phosphate Carbonate Na.CO.,10H.O I, 2, 3, 6
NaNH,HPO,,4H,O 2, 3, 4, 5, 6 Carbonate NasCO"H!O 2,4
Arsenate NasHAsO" 7H.O 2, 3, 4, 5, 6 Cw:bonate, anhydrous
Arsenite NaAsO. 2, 3, 4, 5, 6 Na.CO. 1,2,3,'4,5,6
Bicarbonate NaHCO, 1, 2, 3, 4, 5, 6 Chlorate NaCIO. 3,4,5,6
Bichromate Chloride NaCI I, 2, 3, 4, 5, 6
NaaCr.07,2H.O 2, 3·,4*, 5, 6 Chromate Na.CrO.,4H sO 2,3,4,5,6
Binoxalate NaHC,O"H.O 4 Citrate
Bisulphate, anhydrous 2Na.C.H.07,11 H.O 1,2,3,4,5,6
NaHSO, I, 2, 3, 4, 5, 6 Cobaltinitrite
Bisulphate, cryst., NaaCo(NO.). I, 2, 3, 4, 5. 6
NaHSO"HaO 2,3,4,5 Cyanide NaCN 2,3,4,5,6
Bisulphite NaHSO. 2,4,5,6 Fluoride NaF 2,3,4,5,6
Appendix 805
Sodinm-continued
Ferrocyanide Phosphate, tertiary

Na,Fe(CN)G,lOH,O 4,5 Na.PO,,12H.0 2,3,4,6,6
Formate HCOONa 2,3,4,5,6 Pyrophosphate
Hydroxide NaOH I, 2, 3, 4, 5, 6 Na,p.o"IOH.O 2,3,4-,5.6
Hypophosphite Salicylate Nac,H,O. 2,3,4,5,6
NaH.PO.,H,O 4,6 Silicate Na,SiO.,9H.o 4,
Iodate NaIO. 2,3,4,5,6 Stannate NaSnO,,3H.0 4
IodideNaI 2,3,4,5,6 Sulphate Na.SO.,IOH,O I, 2, 3, 4, 5, 6
Metabisulphite NaS.O, 3 Sulphate, anhydrous
Metaphosphate NaPO. a,4 Na,SO, 1,2,3,4,5,6
MolybdateNa.MoO,,2H,0 2, 3, 4, 5, 6 Sulphide Na. S, 9H.0 I, 2, 3, 4, Ii, 6
Nitrate NaNO. I. 2, 3, 4, 5, 6 Sulphite Na,SO., 7H I O 3,4,6
Nitrite NaNO. I, 2, 3.4,5, 6 Sulphite, anhydrous
Nitroprusside Na.SO. 2,3,4,6,6
Na,Fe(CN)~O,H.O 1,2,3,4,5,6 Tartrate Na.C,H,O.,2H.0 2, 3, 4, 5, 6
Oxalate Na, 0, I, 2, 3, 4, 5, 6 Tetroxalate
Perborate NaBO.,4H.O 2,3,4 NaHC.O"Na.~0.,2H.0 4,5
Peroxide,85% + NaIO. I, 2, 3, 4, 5, 6 Thiocyanate Na NS 2,3,4,5
Phosphate, primary Thiosulphate
!-JaR,PO"H,O . 2,3,4,5,6 Na,S.0.,5H.0 1,2, 3. 4, 5, 6
Phosphate, secondary, cryst., Thiosulphate anhydrous
Na,HPO,,12H,0 1,2,3,4,5,6 Na,S.O. 2,3.4
Phosphate, secondary. Tungstate N a.WO ,,2H,0 I, 2, 3, 4, 5, 6
anhydrous Na,RPO, 1,2,3,4,6
Tin
MetalSn I, 2, 3, 4, 5, 6
Stannous compounds Stannic compounds
Chloride SnC~,2H.0 1. 2, 3, 4, 5, 6 Chloride SnCI,,5H,0 2,3.4,5,6
OxideSnO ' 3,4,5 Chloride, anhydrous
Sulphide SnS 3 SnCI, 3,4
Oxide8nO. 3,4,5
-
Strontium
Acetate Hydroxide Sr(OH).,8H.0 4, 5

2Sr{C.H.0.).,H.O 2,3,4,5,6 Nitrate Sr(NO.). 2, 3, 4, 5, 6
Bromide SrBr.,6H,O 4,5 Osalate SIC.O"HP 4,5
Carbonate SICO, 2,3,4,5,6 Phosphate 4
Chloride SrCI.,6H.0 2,3,4,5.6 Sulphate SISO, 3,4,5,6
Fluoride SrF. 3
Silicot1JllgBtic acid SiO., 12WO., 26H.0 3,4,6

Sulphuric acid, 95-98% HISO, I, 2, 3, 4, 5, 6
Snlphuric acid, fuming, 15% SO. 3; 4, 5, 6
Thorium nitrate Th{NO.),,12H.o 3,4,5 ,
Tnngstic acid I)
Uranyl acetate UO.(C,H.0.l.,2H.O '1,2,3,.,5,6
Uranyl nitrate UO.(NO.).,6H.0 1,2,3,.,5,6
Zinc
\ Metal Zn I, 2, 3, 4, 5,,6 Nitrate Zn(NO.)I,6H 1 O 2,3,4,5,6
Acetate Oxalate ZnC.O.,2HsO 4
Zn(C.H.O.).,2H.O 2,3,4,5.6 OxideZnO 1,2,3,4,5,11
Bromide ZnBr, 4,5 Sulphate ZnSO.,7H1O 1,2.3.4,5,6
Carbonate, basic 2,3,4,5,6 Sulphate, dry
Chromate ZnSO.,H.O 3,4
Chloride, anhydrous
ZnCI.
"2.3,4.5.6 Sulphide, ZnS 5
ORGANIC COMPOUNDS
._--
Acetamide 3.5,6 Amyl acetate 1,2
Acetic anhydride
Acetic acid, glacial
- 1.2,3.4,5.6
1. 2, 3, 4, 5. 6
Amyl alcohol, primary,
normal 2
Acetone 1,2,3,.4,5,6 iso-Amyl alcohol 2,3,4,5
Acetonitrile 6 iso-Amyl nitrite 1,5
Acetyl bromide 1 Aniline 1.2, 3, 4, 5,6
Acetyl chloride 1,3,5,6 Aniline hydrochloride 3,4,5
dl-Alanine 5 Aniline nitrate 5
Allyl alcohol 5 Aniline sulphate 3,4,5
Aminoacetic acid I, 5 I-Arabinose 5
;Benzaldehyde 5 Benzyl alcohol 5

Benzene 1.5,6 Benzyl benzoate 5
Benzidine 6 Benzyl chloride 2,3,4,5
Benzidine acetate 5 Bromobenzene / 6
Benzidine hydrochloride 5 Bromo-cresol purple 6
Benzidine sulphate 5 iso-Butyl acetate 5
Benzil 5 n-Butyl acetate 5
Benzoic acid I, 2, 3, 4, 5, 6 iSd1Butyi alcohol' 5
Benzophenone 5 n-Butyl alcohol 2,3, -i, 5
Benzoyl chloride 1,2,3,4,5 n-ButyTic acid 2
Carbon disulphide 1. 2, 3. 4, 5, 6 Cinnamic acid 5

Carbon tetrachloride I, 2, 3, 4, 5, 6 Citric acid 1,2,3,4,5,6
Carminic acid H Congo red II
Chloroform 1. 2, 3. 4. 5, Ii Cupferron 5
Dextrose 6 Dimethyl yellow 1

Dimethylaminobenzene 1 Diphenylamine 1.5.6
p-Dimethylamino- Diphenylamine hydro-
benzaldehyde 1,5 chloride 5
Dimethylaniline 1,5 Dulcitol
Dimethylglyoxime 1,2,3,5,6
Ether, anhydrous I, 2, 3, 4, 5, (j Ethyl benzoate 5

Ethyl acetate 2,:1,4.5 Ethyl bromide 2,3,5,6
Ethyl acetoacetate 5, 6 Ethylene bromide 5,6
Ethyl alcohol, 90%. Ethyl chlorcarbonate 6
98-100% 1 Ethyl iodide 2,3
Formaldehyde solution, 'ca. 37% l?y weight 1,2,3,4,5.6

Formic acid, sp. gr., 1·20, 87-110% I, 2, 3, 4. 5, 6
Gallic acid
Glycerine
5. 6
I, 2, 3, 4, 5, 6
I Guanidine carbonate 1
Appendix 807
ORGANIC COMPOUNDS-continued
Hippuric acid 5 Ip-HYdroxyphenylarsonic

acid 2
Isatin 5
Lactose 1,2,3,4,5 Lactic acid, 85-88%' 1,2,3,4,5,.6
Maltose 0 Methyl alcohol 1,2,3,4,5

d-Mannitol (mannite) Ii Methyl benzoate Ii
d-Mannose 5 Methyl ethyl ketone 6
Meta nitraniline 5 Methylene blue 6
Meta phenylene diamine Methyl iodide 2,3
hydrochloTide 5 Methyl orange 1,6
Meta toluyJene diamine Methyl red I
hydrochloride 0 Monochloroacetic acid 1, 2, 3, 4, 0, 6
Naphthalene 3,4 Nitrobenzene I

a-Naphthol I p-Nitrobenzoyl chloride 6
,a-Naphthol I p-Nitrochlorbenzene 6
a-Naphthylamine I Nitroso-fJ-naphthol 6
a-NaphthyJamine
hydrochloride 4
bxalic acid, cryst. 1,2,3,4,5,6 Oxinl' 6

Oxalic acid, anhydrous 3,4,5
Petroleum ether (various Phthalic anhydride 2,3.4,5.6

b.p. ranges) 1,2,3,4, Ii Picramic acid 5
Phenol I, 2, 3, 4, 5, 6 Picric acid 1, 2, 3, 4, 5, 6
Phenol red tI n-Propyl alcohol 0
Phenolphthalein 1,2.3,4 iso-Propyl alcohol 5
Phen ylhydrazine Pyridine 1,2, 3, 4, 5, 6
hydrochloride 1,3.4, fi Pyrocatechol 3
Phloroglucinol 1,3,5,6 Pyrogallic acid 2,3,4,5,6
Quinhydrone 1,3,5,6 Quinoline 5
Raffinose hydrate (, R()solicacid 6

Resorcinol 1,2,3,5,6
Saccharin 4 Sucrose I, 2, 3, 4, 5, 6
Salicylic acid 1, 2, 3, 4, 5, 6 Sulphanilic acid 1, 2, 3, 4, 5, 6
Starch, soluble 1,3,4,0,6 Sulphosalicylie acid. 1,6
Succinic acid 1,2,3,4,5, (j
Tartaric acid 1. 2, 3, 4, 5, 6 Toluene 2,3.4,5.11
Tannic acid 2,3,4,5,6 Trichloroacetic acid 1,2,3,4,5, Ii
Thiourea 5 Thymol blue 6
Thymol Ii
Urea 2,3.4,5,11 Uric acid 5
Vanillin I iso-Valerie acid 5
Xylene 2,3,4,5,6
I
A,10. BUFFER SOLUTIONS
1. Buffers suitable for precision standardisation of pll
measurements (D. A. MacInnes, 1939)
lZ0C 2S0C 3S0C
\
pH pH pH
. CH.COOH (O·OIN). CH.COONa
(O'OIN) 4·710 4·700 4·710
CH.COOH (O'lN), CHaCOONa
(0· IN) 4·650 4·640 4:'635
Potassium acid phthalate (0·05M) 4·000 4·000 4·015
I
2. pH Range: 0·65 to 5·20 i 18°C (Walpole, 1914)
50 Ml. N-Sodium acetate +X mi. N-HCl made up to 250 mi.
X (mI.) X (mI.) X (mI.)

pH HCl pH HCI pH HCI
,
0·65 100 1·99 52·5 3·79 42·5
0·75 90 2·32 51·0 3·95 40·0
I
0·91 80 2·64 50·0 4·19 35·0
1-09 70 2·72 49·75 4·39 30·0
1·24 65 3·09 48'5 4'58 25·0
1-42 60 3·29 47·5 4·76 20·0
1·71 55 3·49 46·25 4·92 15·0
1·85 53·5 3·61 45·0 5·20 10·0
3. pH Range: 3·42 to 5·89; 18°C (Walpole, 1914)
10 Ml. mixtures of X ml. of O·ZN acetic acid and Y ml. of

O'ZN sodium acetate.
X (mI.) Y (mi.)' pH X (mI.) Y (ml.) pH

-
9'5 0·5 3·42 4·0 6·0 4·80
9·0 1·0 3·72 3·0 7·0 4·99
8·0 2·0 4·05 2·0 8-0 5·23
7·0 3-0 4·27 1-5 8-5 5·37
6-0 4·0 4·45 1·0 9-0 5-57
5·0 5·0 4·63 0·5 9·5 5'89
Appendix 809
4. pH Range: 2·2 to 8·0 (McIlvaine, 1921}
20 MI. mixtures of X mi. of 0'2M Na 2 flPO. and Y mi. of O'lM

citric acid.
X (mI.) Y(m!.) X(m!.) Y (ml.)
Na,HPO. Citric acid pH Na 2 HPO. Citric acid pH
-
0·40 19·60 2·2 10·72 9·28 5·2
1·24 18·76 2·4 11·15 8·85 5·4
2·18 17·82 2·6 1l·'()0 8·40 5·6
3·17 16'83 2·8 12·09 7·91 5·8
4·11 15'89 3·0 12·63 7·37 6·0
4·94 15·06 3·2 13·22 6·78 6·2
5·70 14'30 3'4 ]3·85 6·15 IN
6·44 13'56. 3·6 14·55 5·45 6·6
7·10 12·90 3·8 15·45 4·55 6·S
7·71 12·29 4·0 16·47 3·53 7·0
R·28 11·72 4·2 17·39 2·61 7·2
8·82 11-18 4·4 18·17 1·83 7·4
9·35 10·65 4·6 18·73 1·27 7·6
9·86 10·14 4·8 19·15 0·85 7·8
I
10·30 9'70 5·0 19·45 0·55 8·0
5. pH Range: 5·29 to 8·04 ; l8°e (Sorensen, 1909-12)
10 MI. mixtures of X mi. MI15 Na.HPO. and Y mt. of MilS

KH.PO •.
X (mI.)
NaHPO.
Y (mI.)
KH,PO. pH I NaIHPO.
X (mI.) Y (mI.)
KH.PO. pH
0·25 9·75 5·29 5·0 5·0 6'81

0·5 9·5 5'59 6·0 4·0 6·98
1·0 9·0 5·91 7·0 3·0 7·17
2·0 8·0 6·24 8·0 2·0 7·38
3·0 7'0 6'47 9·0 1'0 7·73
4·0 6·0 6·64 9·5 0·5 8·04
DO·
810 Quantitative Inorganic A nal;Jsis
6. pH Ranges: 2·2 to 3·8, 4·0 to 6·2, 5·8 to 8'0, 1·8 to 10·0 ;
20°C (Clark and Lltbs, 1916)
-
(A) pH 2·2-3·8. 50 MI. 0·2M J": H Phthalate + P mi. 0·2M HCI,
diluted to 200 mi.
(fl) pH .. ·0-6·2. 60 MI. 0'2M K H Phthalate + Q mi. O·2M NaOH,
diluted to 200 mi.
(C) pH 5,8-8,0. 50 MI. 0·2M KH.PO. + R mi. 0·2 M NaOH,
diluted to 200 mi .
. (fl) pH 7'8-10,0. 50 MI. 0·2M H.BO. atld 0'2M RCI· + S mI.
O·2M NaOH, diluted to 200 mi.'
A B
I C D
[' (m1.) Q (mI.) R (mI.) I pH S (mI.)
HCI
-
pH
-NaOH__ .
pH NaOH NaOH pH
4fi·1l0 2·2 0·40 4·0 3·66 5·8 2·65 7·8

39·60 2·4 3·65. 4·2 5·64 6·0 4·00 8·0
33·00 2·6 7·35 4·4 8·55 6·2 5·90 8·2
26·50 2·8 12·00 4·6 12·60 6·4 8·55 8·4
20·40 3·0 17·50 4·8 17·74 6·6 12·00 8·6
14'8'0 3·2 23·65 5·0 23·60 6·g 16·40 g'8
9·95 3·4 29·75 5·2 29·54 7·0 21-4·0 9·0
6·00 3·6 35·25 5·4 34·90 7·2 26·70 9·2
2·65 3·8 39·70 5·6 39·34 7·4 32·00 9·4
- - 43·10 5·8 42.74· 7·6 36·85 9·6
- - 45·40 6·0 45·17 7·8 40·80 9·8
- - 47·00 6·2 46·85 8-0 43·90 10·0
7. pH Range: 6·80 to- 9·60; 25°C (Michaelis, 1930)
10 MI. mixtures of X mI. of O'IM sodium diethylbarbiturate and

Y mI. of O'lM HCI.
X (mi.) Y (mI.) X (mI.) Y (mI.)

Na salt HCI pH Na salt HCl pH
.- . -
5·22 4·78 6·80 8·23 1·77 8·40
5·36 4·64 7·00 8·71 1·29 8·60
5·54 4·46 7·20 9·08 0·92 8·80
5·81 4·19 7·40 9·36 0'64 9·00
6·15 3'85 7·60 9·52 0'48 9·20
6·62 3·38 7·80 9·74 0·26 9·40
7·16 2·84 8·00 9'85 0·15 9·60
I
• Thatis, a solution containing 12·:l1:i9 g. of H.BO, and 14'911 g. of KCI per

litre.
Appendix 811
8. pH Range: 8·45 to 12·11; 24°C (S6rensen-Walbum, 1920)
10 MI. mixtures of X mI. of O'lM glycine + O'lM NaGI· and

Y mI. ofO·1M NaOH.
---
X (ml.) X (ml.)
Glycine Y _(ml.) pH Glycine Y (ml.) pH
-NaCl NaOH II -NaCl NaOH
9-5 0-5 8-45 5·0 5·0 11·14

9·0 1·0 8·79 4·9 5·1 11·39
8·0 2·0 9·22 4·5 5·5 11·92
7-0 3-0 9-56 4-0 (i·O 12·21
6·0 4·0 9·98 3·0 7·() 12·48
5·5 4·5 10·32 2·0 8·0 12·6ti
5·1 4-9 10-90 1·0 9·0 12·77
9. pH Ranges: 1·40 to 2,20, 2·30 to 3·30, 2·90 to 4·10,

3·40 to 5,20, 4·80 to 6·2p; 25°C (German and Vogel, 1937)
X MI. of O·lM para-toluenesulphonic acid monohydrate (19'012 g.

per litre) and Y mI. of O'lM sodium para-tolumesulphonate
(111-406 g. per litre), diluted to 100·0 mi.
X (ml.) Y(ml.) pH X (mJ.) Y(ml.) pH

48·9 1-1 1·40 36·8 13·2 1·90
37·2 12·8. 1·50 40·0 10·0 2·00
27'4 22·6 1·60 42·4 7·6 2·10
19·0 31·0 1·70 45·6 4·4 2·20
16·6 33·4 1·80
I --
X MI. of O·OlM para-toluenesulphonic acid monohydrate (1·9012

g. per litre) and Y mi. of 0-01M sodium para-toluenesulphonate
(1·9046 g. per litre), diluted to 100·0 mI.
X (ml.) Y (ml.) pH I X (mI.) Y (mI.) pH

50·0 0·0 2·30 12·4 37·6 2·90
44·2 5·8 2·40 10·0 40·0 3·00
35·8 14·2 2·50 8·0 42·0 3·10
27·6 22·4 I
I 2·60 6·2 43·8 3·20
21·2
18·5
28'8
31·5
I 2·70
2·80
5·0 45·0 3·30
I
1
• That is, a solution containing 7'505 g. of glycerine + 5·85 g. of NaCI

per litre.
X MI. of O·OIMfuroic acid (1·1203 g. per litre) and Y mi. of O'OIM

sodium furoate (1'3403 g. per litre) diluted to 100·0 mt
X (mI.) Y (mi.) pH II X (tni.) Y (ml.) pH

45·2 4·8 2·90 14·8 35·2 3·60
39·6 10·4 3·00 12·3 37·7 3·70
34·0 16·0 3·10 10·1 39·9 3·80
28·4 21·6 3·20 8·4 41·6 3·90
24·8 25·2 3·30 6·7 43·3 4·00
21'4 28·6 3·40 5·4 44·6 4·10
IS·O 32·0 3·50
X. MI. of O'OIM phenylacetic acid (1'3606 g. per litre) and Y mi.

of O·OIM sodium phenylacetate (1·5806 g. per litre). diluted to
100·0 mI.
X (mI.) Y (mi.) pH X (mI.) Y (mI.) pH
48·7 1·3 3·40 22·0 28·0 4·40
46·6 3·4 3·50 19·4 30·6 4·50
44·4 5·6 3·60 16·9 33·1 4·60
4'2·2 7·8 3·70 14·4 35·6 4·70
39·9 10·1 3·80 12·1 37·9 4·80
37·2 12·6 3·90 10·2 39·8 4·110
34·4 15·6 4·00 8·6 41·4 5·00
30·8 19·2 4·10 7·2 42·8 5·10
28·0 22·0 4·20 5·7 44·3 5·20
25·1 24·9 4·30
X MI. of O'OIM sodium hydrogen succinate (0·5903 g. of succinic

acid + 0·8102 g. of disodium succinate per litre) and Y mi. of
O·OIM disodium succinate (1'6203 g. per litre). diluted to 100·0 mI.
X (mI.) Y (mI.) pH
I X (mI.) Y (ml.) pH
48·8 1·2 4·80 21·0 29·0 5·60
46·9 3·1 4·90 17·6 32·4 5·70
44·4 5·6 5·00 14·7 35·3 5·80.
40·4 9·6 5·10 12·0 38·0 5·90
36·2 13·8 5·20 9·9 40·1 6·00
32·6 17·4 5·30 7·8 42·2 6·10
28·6 21·4 5·40 5·6 44·4 6·20
24·7 25·3 5·50
,
Appendix 813
A, 11. COMPARISON OF METRIC AND OTHER UNITS
Metric IIYStem.-The primary unit of length is the metre, which is the
distance between two lines, under standard conditions, engraved on a
prc.totype preserved in the Bureau International des Poids et Mesures,
Sevres, France.
The primary unit of weight is the kilogram, which is the mass of
a prototype preserved in the same place.
The primary unit of volume is the litre. and is the volume occupied by
one kilogram of pure water at 4°C under a pressure of 760 mm. of
mercury (see, however, Section n, 17).
The millilitre (mI.) is the thousandth part of a litre. One millilitre
is nearly, but not absolutely equal to cubic centimetre (c.c.).:
1000 ml. = 1000·027 C.c.
UNITS OF LENGTH
1 millimetre (mm.) = 0·03937 inch. 1 inch (in.) = 2(J-400 millimetres.
1 centimetre (cm.) = 0·3937 inch. 1 inch = 2·5400 centimetres.
1 metre (m.) =39·370 inches. 1 foot (ft.) = 304·801 milijmetre'i.
1 metre = 3·2808 feet. 1 foot = 0·30480 metre.
1 kilometre (km.) = 1093·61 1 yard (yd.) = 0·91440 metre.
yards. 1 mile = 1·60934 metres.
i kilometre = 0·62137 mile.
UNITS OF AREA
1 squaremm. = 0·001550 sq. in. I square in. = 645·16 sq. rom.
1 squlJ.re cm. = 0·1550sq.in. 1 square in. = 6·4516 sq. cm.
1 square metre= 1550'0 sq. ins. 1 square ft. = 929·03 sq. cm.
1 square metre'= 10·7638 sq. f¢. 1 square ft. = 0·9290 sq. m.
1 square mene= 1·1960 sq. yds. 1 square yd. = 0·86313 sq. m.
1 square km. 0·38610 sq.mile. I acre = 0·40469 hectare.
1 hectare 2·4710 acres. 1 square mile = 2·5900 sq. km.
UNITS OF VOLUME
1 cubic mm.=0·OO006102cubicin. 1 cubic·in. = 16,387 cubic mm.
1 cubic cm. (c.c. or mI.) 1 cubic in. = 16·387 mI.
= 0·061024 cubic in. 1 cubic ft. = 0·028317 cubic m.
1 cubic m. = 61,024 cubic ins. 1 cubic ft. = 28·3171itres.
1 cubic m. = 35·3144 cubic ft. I cubic yd. = 0·7645 cubic m.
= 1'3079 cubic yM
'MEASURES OF LIQUID CAPACITY

Imperial or British Units
1 gill = 142·06 ml.
1 pint = 4 gills = 568·25 ml.
1 quart = 2 pints = 1·13649Iitres.
1 gallon = 4 quarts = 4·54596 litres.
Apothecaries' Units (British)
I minim = 0·059192 m!.
1 fluid scruple = M minims = 1·1838 m!.
I fluid drachm = 3 scruples = 3·5515 m!.
I fluid ounce = 8 fluid drachms = 28'412 m!.
1 pint = 20 fluid ounces = 568·25 m!.
1 gallon 8 pints = 4·54596 litres.
U.S. Units.
1 gill _ = 118·292 mI.
1 pint = 4 gills = 473·167 m!.
1 quart = 2 pints = 0·94633 litre.
I gallon = 4 quarts = 3·78533 litres.
Apothecaries. Units (U.S.)

1 minim. = 0·06161: m!.
1 fluid drachm 60 minims = 3·69661 m!.
1 fluid ounce 8 fluid drachms = 29·5729 m!.
I pint 16 fluid o,unces = 473·167 ml.
1 gallon 8 pints = 3·78533 litres.
UNITS OF MASS
Avoirdupois Weight
I grain =-64·799 mg.
1 dram 27 ·3437 grains = 1·77185 g.
I ounce 16 drams = 437·5 grains = 28·34954 g.
I pound 16 ounces = 700(} grains = 453'5926 g.
1 stone = 14 pounds = 6·35030 kg.
I quarter = 28 pounds = 12·70059 kg.
1 hundredweight ll2 pounds = 50·80238 kg.
I gram = ] 5·4325 grains.
I decagram 10 grams = 5·64383 drams.
1 hectogram 100 grams = 3·52739 ounces.
1 kilogram 1000 grams = 2·204621 pounds.
I myriagram 10 kilograms = 22·04621 pounds.
1 quintal 100 kilograms = 1·98641 hundredweights.
1 millier I metric ton = 1000 kilograms = 0·9842059 ton.
Troy Weight
I grain = 64·799 mg.
1 pennyweight (dwt) = 24 grains = 1·29598 g.
1 ounce = 20 pennyweights = 480 grains = 31·1030 g.
I pound 12 ounces = 373·2418 g.
A p'othecaries' Weight'
I grain '== 64·79g mg.
I scruple 20 grains = 1·29598 g.
I drachm = 3 scruples = 3·88794 g.
I ounce = 8 drachms = 480 grains = 31·1035 g.
1 pound = 12 ounces = 373·2418 g.
Appendix 815
A.12. DECIl't'IALS OF AN INCH. I.S.W.G.• AND MILLIMETRES
Decimals Milli- Decimals Milli-

of an inch I.S.W.G. metres of an·inch I.S.W.G. metres
0-400 0000 10·160 0·040 19 1·016

0-372 000 !H49 0·036 20 0·914
0-348 00 8·839 0·032 21 0·813
0-324 0 8·230 0·028 22 0·711
0-300 1 7·620 0·024 23 0·610
0-276 2 7·010 0·022 24 0·559
0-252 3 6-40i 0·020 25 O·50S
0-232 4 5·893 0·018 26 0·457
0·212 5 5·385 0·016 27 0·416
0-192 6, 4·877 0·015 28 0·376
0-176 7 4·470 0·014 29 0·345
0-160 8 4·064 0·012 -I- 30 0·315
0-144 9 3·658 0·012 - 31 0·295
0-128 10 3·251 O·OIl 32 0·274
0·U6 11 2·946 O·OlO 33 0·254
0·104 12 2·642 0·009 34 0·233
0·092 13 2·337 .0·008 -I- 35 0·213
0·080 14 2·032 0·008 - 3U 0·193
0·072 15 1·829 0·007 37 0·172
0·064 16 . 1'626 0·006 38 0·152
0·056 17 1·422 0·005 + 39 0·132
0·048 18 1·219 0·005 - 40 0·122
A.13. THE GREEK ALPHABET
Greek ·Greek English Greek Greek English

Leiter Name Equiva- Leila Name Equiva-
lent lel/t
-
=
A a Alpha a N Nu n
B f3 Beta b E "t Xi x
r 'Y Gamma g 0 0 Omicron 0
LI
E
Z
H TJ
,
0
€
Delta
Epsilon
. Zeta
Eta
d
C
z
e
II
P
.E
T
7T
P
a
T
Pi
Rho
Sigma
Tau
p
r
5
t
e e Theta th Y tJ Upsilon u
I L Iota i ifJ cp Phi ph
K K Kappa k X X Chi eh
A ,\ Lambda 1 lJI rP Psi ps
M Jl. Mu m Q w Omega 0
I i
A,14. SUGGESTED SCHEME OF STUDY FOR ·INTERMEDIATE
SCIENCE AND ENGINEERING, mGHER SCHOOL CERTlFICATE
AND ORDINARY NATIONAL CERTIFICATE EXAMINATIONS
IN CHEMISTRY, AND EXAMINATIONS OF SIMILAR CHARACTER
The following scheme is suggested as suitable for a course of training
in quantitative analysis for beginners. It covers the requirements of
the intermediate examination in science and engineering, the higher
school certificate examination, the first medical examination, the
ordinary national certificate in chemistry, and examinations of similar
scope. Experiments marked with an asterisk are optional.
Acidimetry and Alkalimetry
(i) Preparation of approx. O·IN hydrochloric acid (Section m, 4) and
standardisation agai.nst (a) anhydrous sodium carbonate (method of Note 1 in
Section In, 4A) or (b) borax (method of Note 1 in Section In, 4B).
(ii) Preparation of approx. O·IN sodium hydroxide solution (Section In, 5;
Procedure A) and standardisation against (a) ca. O'lN hydrochloric acid
(Section m, 6; Procedure A) or (b) potassium hydrogen phthalate (Section
m, 6; Procedure B).
(iii) Determination of Na.CO a in washing soda with standard hydrochloric
acid (Section m, 9).
(iv) Determination of strength of glacial acetic acid with standard sodium
hydroxide solution (Section m, 10).
(v) *Determination of H.S0 4 in the concentrated acid with standard sodium
hydroxide solution (Section m, 11).
(vi) Analysis of commercial caustic soda (mixture of hydroxide and car-
bonate (Section m, 14; Procedure B).
(vii) Determination of Na.B.O, in borax (Section m, 17; titration with
standard acid and indicator only).
(viii) Determination of NH3 in ammonium chloride (Section m, 18; indi-
cator method).
Determinations with Potassium Permanganate Solution

(i) Preparation of O'lN potassium permanganate (Section m, 47).
(ii) Standardisation of the potassium 'Permanganate solution (a) with
arsenious oxide. (Section m, 48A; Note 1). (b) with sodium oxalate (Section
m, 48B; Note 1).
(iii) Determination of iron in crystallised ferrous sulphate (Section m, 50).
(iv) Determination of iron in ferric ammonium sulphate (Section In, 52;
reduction with stannous chloride as in Section m, 5lA).
(v) *Determination of total iron in an iron ore (Section m. 53; reduction
with stannous chloride as in Section m, 5lA).
(vi) Determination of HiO, in the commercial solution (Section m, 55).
(vii) Determination of calcium in calcium carbonate (Section m, 54).
(viii) Determination of MnG, in pyrolusite (Section m, 57; Procedure B).
(ix) *Determination of KNO, in commercial potassium nitrite (Section m.
58).
Determinations with Potassium Dichromate Solution
(i) Prt;paration of O·lN potas~ium dichromate (Section m, 64).
(ii) Check standardisation of pota~sium dichromate soltltion against pure
iron (Section m, 65).
(iii) Determination of iron in crystallised ferrous sulphate or ferrous ammon-
ium sulphate (Section m, 66) ; sodium diphenylamine sulphonate or N-pheny-
lanthranilic acid as indicator.
(iv) Determination of iron in ferric ammonium sulphate (Section m, 67).
(v) *Determination of total iron in an iron ore (Section m. 68).
(vi) *Determination of chromium in chrome alum (Section m. 70).
Apflendix 817
Iodimetry and Iodometry
(i) Preparation of approx. O'lN sodium thiosulphate solution (Section m.
101).
(ii) Standardisation of the thiosulphate solution against (a) standard potas-
sium iodate solution (Section m, 102A; Note 1), (b) standard potassium dichro-
mate solution (Section m.l02D).
(iii) Preparation of approx. O·IN iodine solution (Section m. 103; Method
A). .
(iv) Standardisation of the iodine solution against (a) pure arsenious oxide
(Section m. 104A), (b) standard so,lium thiosulphate solution (Section m.
1MB).
(v) Determination of copper in crystallised copper sulphate (Section- m,
105; Method A).
(vi) Determination of MnO. in pyrolusite (Section In, 107).
(vii) Determination of H.O. in the commercial solution (Section m,llO).
(viii) Determination of available chlorine in bleaching powder: (a) iodo-
metric method (Section m. Ill; Procedure A), (b) arsenite method (Section
m. 111; Procedure B).
(ix) Determination of antimony in tartar emetic (Section In, 113).
(xl *Determination of tin in crystallised stannous chloride (Section m.
116; Procedure A).
(xi) *Determination of purity of crystallised sodium sulphite (Section m,
118).
(xii) *Determination of HIS in aqueous solution (Section m. 119).
Precipitation Reactions
Use ot Standard Solutions ot Silver .Nitrate and 01 Ammouium Thiocyanate
(i) Preparation of approx. O'lN silver nitrate solution (Section m, 22 ;
Procedure B).
(iil Standardisation of the silver nitrate solution against standard sodium
chloride solution (al with potassium chromate as indicator-Mohr's method
(Section m, 23A). (b) with fluorescein as indicator-Fajan's ~ethod (Section
m,23B).
(iii) *Analysis of a mixture of potassium chloride and bromide (Section
lII,aO).
(iv) Preparation of approx. O-lN sodium thiosulphate solution (Section m,
31).
(v) Standardisation of the ammonium thiocyanate solution against standard
silver nitrate solution by Volhard's method (Section m. 31).
(vi) *Determination of silver in a silver alloy (Section m, 32).
(vii) Determination of the Hel content of concentrated hydrochloric acid
(Section m. 33).
(viii) *Determination of KeN in commercial potassium cyanide (Section
m. 41; Procedure A).
Simple Gravimetric Analysis
(i) *Determination of water of crystallisation in crystallised barium
chloride (Section IV, 4).
(ii) *Determination of chlorine in sodium chloride (Section IV, 5).
(iii) *Determination of sulphate in potassium sulphate (Section IV, 6).
(iv) *Determination of iron in ferrous ammonium sulphate (S"ection IV. 7).
A. 15. SUGGESTED SCHEME OF STUDY FOR B.Sc. SPECIAL

CHEMISTRY, HONOURS B.Sc. CHEMISTRY, ASSOCIATESHIP
OF THE INS'tnoTE OF CHEMISTRY, AND EXAMINATIONS
OF SIMILAR SCOPE
The scheme given below is proposed as a basis for a course in quanti-
tative analysis for advanced courses in chemistry. The experiments
marked with an asterisk are of more difficult character; they are
suitable for A.I.e. students and may be omitted, if desired, by B.Sc.
honours students. It is not expected that anyone student will be able
to carry out all the determinations listed below, but a sufficient number
is given to allow a choice at the discretion of the teacher. It need
hardly be stated that many teachers will prefer to adopt their _own
courses, and it is believed that details of most of the necessary deter-
minations will be found in the book.
VOLUMETRIC ANALYSIS
Acidimetry and Alkalimetry
(i) Calibration of weights (Section n. 8).
(ii) Calibration of volumetric apparatus:
(a) volumetric flasks, 250 mI., 500 mI., 1000 mI. (Section n. 20) ;
(b) pipettes, 10 ml., 25 mt, 50 ml. (Section n. 21) ;
(e) burette (Section n. 22).
(iii) Preparation of O'IN hydrocbloric acid (Section m.3 or nI, 4). Stan-
dardisation against:
(a) borax (Section m. 4B) ;
(b) anhydrous sodium carbonate (Section m. 4A).
(iv) Determination of S03 in oleum (Section m. 12).
(v) Determination of NH3 in an ammonium salt (Section m. 18; direct
method) or of NaNO. in Chili saltpetre (Section m. 19).
(vi) Determination of nitrogen by Kjeldahl's method (Section m.20).
Precipitation and Complex-Formation Reactions

(i) Preparation and standardisation of O·IN silves-nitrate (Sections 22 m.
and m, 23).
(ii) *Determination of chloride and iodide or of bromide and iodide in admix-
ture (Section m. 29).
(iii) Preparation and standardisation of o·} N ammonium thiocyanate
(Section m. 31).
(iv) Analysis of a_ hyposulphite (Section m. 36).
(v) Determination of cyanide in potassium cyanide (Section m. 41).
(VI) Determination of nickel:
(a) with pyridine and staflliard ammonium thiocyanate (Section m. 38);
(b) with potassium cyanide solution (Section m.42).
Oxidations with Potassium Permanganate

(i) Preparation of 0·1 N potassium permanganate and standardisation
against:
'(a) arsenious oxide (Sectionm. 48A) ;
(b) sodium oxalate (Section m. 48B).
(ii) Determination of total iron in an iron ore (Section m, 53; reduction
with (a) stannous chloride and (b) the Jones' reciuctor-Section m, 51).
(iii) Analysis of sodium peroxide (Section m, 56).
(iv) Determination of MnO. in pyrolusite (Section m. 57).
(v) Determination of a nitrite (Section m, 58).
(vi) Determination of a persulphate (Section m. 59).
(vii) Determination of manganese in steel by the bismuthate method
(Section m. 60).
Oxidations with Potassium Dichromate

(!1 Preparation of o·IN potassium dichromate (Section m. 64).
(11) Determination of ferrous and ferric iron in an iron ore (Section m 69).
liii) Determination of chromium in chromite (Section m. 71). •
Appendix 819
Oxidations with Cerie Sulphate
(i) Preparation of approx. O'lN ceric sulphate solution (Section m, 74)
and standardisation against :
(a) arsenious oxide (Section m, 75A) ;
(b) pure iron (Section m, 75B). .
m.
(ii) Determination of an oxalate (Section 76).
(iii) Determination of iron in an iron ore (Section m, 77).
(iv) Determination of a ferrocyanide (Section m, 78).
(v) Determination of a nitrite (Section m, 79).
Reductions with Titanous' Sulphate
(i) *Preparation of titanous sulphate solution (Section m, 87B) and stan-
dardisation against pure iron (Section m, 88A).
(ii) *Determination of iron in an iron ore (Section m, 89).
IODIMETRY AND IODOMETRY

Dllterminations with Sodium Thioswphate
(i) Preparation of O'lN sodium thiosulphate (Section Ill, 101) and stan-
dardisation against:
(a) potassium iodate (Section m, 102A) ;
(b) potassium dichromate (Section m, 102D) ;
(c) pure copper (Section m. 102E).
(ii) Preparation of O'lN iodine (Section m, 103) and standardisation.
against: \
(a) arsenious oxide (Section m, l04A) ;
(b) O'IN sodium thiosulphate (Section m, l04B).
(iii) Determination of copper in an ore (Section m. 106).
(iv) Determination of manganese dioxide in pyrolusite (SectionIn.107).
(v) Determination of available chlorine in bleaching powder (Section IU.
Ill).
(vi) Determination of a chlorate or a bromate (Section m.108).
(vii) *Determination of tin in an alloy (Section m, 116A).
(viii) *Det~rmination of antimony and tin in type metal (Section nI, 117).
(ix) Determination of a sulphite (Section m, 118). .
Oxidations with Potassium Iodate
(i) Preparation of O'IN (or O·025M) potassium'iodate (Section m. 123).
_Iii) Determination of an iodide (Section m, 124). .
(iii) Determination of arsenic or of antimony (Section m, 125).
(iv) Determination of copper (Section m, 126).
(v) *Determination of tin (Section m, 128).
(vi) Determination of hydrogen peroxide (Section m, 129).
Oxidations with Potassium Bromate

(i) Preparation of O·lN potassium bromate (Section'm, 136).
(ii) Determination of antimony or of arsenic (Section m, 137).
(iii) Determination of aluminium or of magnesium by the oxine method
(Section In, 136; Procedure A and B).
GRAVIMETRIC ANALYSIS
(i) Determination of chloride as AgCI (Section IV, 5).
(ii) Determination of sulphate as BaS0 4 (Section IV, 6).
(iii) *Determination of sulphur in iron pyrites (Section IV, 7).
(v) Determination of iron as Fep. (Section IV. 8).
(v) Determination of aluminium as Alps (Section IV, 9) ..
(vi) Determination of calcium as CaCOa or as CaO (Section IV,10).
(vii) Determination of magnesium as Mg.P.O, (Section IV, 11).
(viii) Determination of nickel with dimethylglyoxime (Section IV. 12).
(ix) Determination of lead as PbCrO, or as PbSO~ (Section IV, 14).
(x) Determination of mercury as HgS or as Hg[Zn(CNS),] (Section lV, 16).
(xi) -Determination of bismuth as BiOI (Section IV, 17).
(xii) -Determination of cadmium by the pyridine method (Section IV, 18).
(xiii) Determination of copper as Cu.(CNS). (Section IV. 19).
(xiv) -Determination of tin as SnO. (Section IV, 22). ,
(xv) Determination of aluminium as the 8-hydroxy-quinolate (Section lV,
28).
(xvi) Determination of chtomium by the cyanate method or by the mer-
curous chromate method (Section IV, 30).
(xvii) Determination of iron with cupferron (Section lV, 31).
(xviii) Determination of cobalt with a-nitroso-,B-naphthol (Section IV, 33).
(xix) Determination of zinc as ZnNH,PO, and lor Zn.P,O, (Section IV, 34).
(xx) Determination of sodium as sodium zinc uranyl acetate (Section
IV,47).
(xxi) Determination of potassium as K.Na[Co(NO.).J (Section IV. 48).
(xxii) "Determination of fluorine as PbCIF (Section IV. 57).
(xxiii) Determination of phosphorus as Mg.P.O. (Section IV, ~).
(xxiv) Determination of silica in a soluble silicate and in an insoluble silicate
(Section IV. 70).
(xxv) Determination of a carbonate (Section IV, 78).
Electrolytic Determinations
(i) Determination of copper (Section IV, 18).
(ii) Determination of lead (Section IV, 79).
(iii) Determination of nickel (Section IV, 81).
(iv)· Determination of zinc (Section IV, 84).
Simple Gravimetric Separationa
(i) Iron and aluminium (Section IV, 86A).
(ii) Iron and chromium (Section IV, 86B).
(iii) Nickel and zinc (Section IV, S6D).
(iv) Calcium and magnesium (Section lV, 86E)
(v) Calcium and barium (Section IV, 86F).
(vi) Sodium and potassium (Section IV, 86, I).
Analysis of Complex Materials

(i) .Brass (Section IV, 88).
(ii) -Solder (Section lV, 90).
(iii) -German silver (nickel-silver) (Section IV, 89).
(iv) -Silver coinage alloy (Section IV, 91).
(v) -Limestone or a dolomite (Section lV, 94).
(vi) -Felspar (Section IV, 95).
(vii) .Portland cement (Section lV, 96).
COLORIMETRIC ANALYSIS
(i) .Determination of ammonia with Nessler's reagent (Section V, 8).
(ii) -Determination of iron by the thiocyanate method (Section V, lOA).
(iii) .Determination of manganese as permanganic acid (Section V, n).
(iv) ·Determination of titanium by the hydrogen peroxide method (Section
V,12).
(v) ·Determination of chromium with diphenylcarbazide (Section V, 14).
(vi) .Determination of lead by the dithizone method (Section V, 22).
(vii) -Determination of copper by the ferrocyanide method andlor the
sodium diethyldithiocarbamate method (Section V, 23). .
(viii) -Determination of silicate by the molybdate method (Section V, 26).
Appendix 821
GAS ANALYSIS AND GAS- VOLUMETRIC ANALYSIS
(i) ·Determination of oxygen in air (Section VI, 9).
(ii) "'Analysis of coal gas with the Bunte burette (Section VI, 6B).
(iii) • Analysis of coal gas with the Orsat apparatus (Section VI, 6C).
(iv) .Analysis of coal gas with the Ambler apparatus (Section VI, 6D).
(v) .Determination of a nitrate with the Lunge nitrometer (Section VI, 11).
(vi) ·Evaluation of pyrolusite (Section VI, 13).
(vii) *Determination of available chlorine in bleaching powder (Section VI,
14).
(viii) .Determination of ammonia in an ammonium salt (Section VI. 16).
A.16: FOUR-FIGURE LOGARITHMS
II0
10 0000
1 I 213
0043 0086 0128
! 41
0170
5
0212
6 7 8
0253 0294 0334 0374
I9
4
MEAN DIFFERENCES
11213 41 5 16 71 8 9
I
812 17 21 25 29 3337
11 0414 0453 0492 0531 0569 0607 0645 0682 0719 0755 4 811 15 19 23 26 3034
12 0792 0828 0864 0899 0934 0969 1004 1038 1072 1106 3 7 10 14 17 21 24 2831
13 1139 1173 1206 1239 1271 1303 1335 1367 1399 1430 3 () 10 13 16 19 23 2629
14 1461 1492 1523 1553 1584 1614 1644 1673 1703 1732 3 6 9 12 15 18 21 2427
15 1761 1790 1818 1847 1875 1903 1931 1959 1987 2014 3 6 811 14 17 20 2225
16 2041 2068 2095 2122 2148 2175 2201 2227 2253 2279 3 5 811 13 16 18 2124
17 2304 2330 2355 2380 2405 2430 2455 2480 2504 2529 2 5 710 12 Hi 17 2022
18 2553 2577 2601 2{)25 2648 2672 2695 2718 2742 2765 2 5 7 9 12 14 16 1921
19 2788 2810 283~ 2856 2878 2900 2923 2945 2967 2989 2 4 7 9 11 13 16 1820
20 3010 3032 3054 3075 3096 3118 3139 3160 3181 3201 2 4. 6 8 11 13 15 17 19
21 3222 3243 3263 3284 3304 3324 3345 3365 33853404 2 4 6 8 10 12 14 1618
22 3424 3444 3464 3483 3502 3522 3541 3560 3579 3598 2 4 6 810 12 14 15 17
23 3617 3636 3655 3674 3692 3711 3729 3747 3766 3784 2 4 () 7 9 11 13 t5 17
24 3802 3820 3838 3856 3874 3892 3909 3927 3945 3962 2 4 5 7 9 11 12 1416
25 3979 3997 4014 4031 4048 4065 4082 4099 4116 4133 2 3 5 7 9 10 12 14 15
26 4150 4166 4183 4200 4216 4232 4249 4265 4281 4298 2 3 5 7 8 10 11 13 15 .
27 4314 4330 4346 4362 4378 4393 4409 4425 4440 4456 2 3 I) 6 8 911 1314
28 4472 4487 4502 4518 4533 404,8 45{)4 4579 4594 4609 2 3 5 () 8 911 12 14
29 4624 4639 4654 4669 4G83 4698 4713 4728 4742 4757 1 3 4 6 7 910 12 13
30 4771 4786 4800 4814 4829 4843 4857 4871 48g6 4900 1 3 4 6 7 910 11 13
31 4914 4928 4942 4955 4960 4!lSa 4997 5011 5024 5038 1 3 4 6 7 810 11 12
32 5051 5065 5079 5092 5105 5119 5132 5145 5159 5172 1 3 4 5 7 8 9 11 12
33 5185 5198 52115224 5237 5250 5263 5276 5289 5302 1 3 4 5 6 8 9 10 12
34 5315 5328 5340 5353 5366 5378 5391 5403 5416 5428 1 3 4 5 6 8 9 10 11
35 5441 5453 5465 5478 5490 5502 5514 5527 5539 5551 1 2 4 5 6 7 9 10 II
36 5563 5575 5587 5599 5611 5623 5635 5647 5658 5670 1 2 4 5 6 7 810 11
37 5682 5694 5705 5717 5729 5740 5752 5763 5775 5786 1 2 3 5 6 7 8 9 10
38 5798 5809 5821 5832 5843 5855 5866 5877 5888 5899 1 2 3 5 6 7 8 910
39 5911 5922 5933 5944 5955 59{){) 5977 5988 5999 6010 1 2 3 4 5 7 . 8 910
40 6021 6031 6042 6053 6064 6075 6085 6096 6107 6Il7 1 2 3 4 5 6 8 910
41 6128 6138 6149 6160 6170 6180 6191 6201 6212 6222 1 2 3 4 5 6 7 8 9
42 6232 6243 6253 6263 6274 6284 6294 6304 6314 6325 1 2 3 4 5 6 7 8 9
43 6335 6345 6355 6365 6375 6385 6395 6405 6415 6425 1 2 3 4 5 6 7 8 9
44 6435 6444 6454 6464 6474 6484 6493 6503 6513 6522 1 2 3 4 5 6 7 8 9
46 6532 6542 6551 6561 6571 6580 6590 6599 6609 6618 1 2 3 4 5 6 7 8 9
46 6628 6637 6646 6656 6665 6675 6684 6693 6702 6712 1 2 3 4 5 6 7 7 8
47 6721 6730 6739 6749 6758 6767 6776 6785 6794 6803 1 2 3 4 5 5 6 7 8
48 6312 6821 6830 6848 6848 6857 6866 6875 6884 6893 1 2 3 4 5 5 6 7 8
49 6902 {)911 6920 6928 6937 6946 6955 6964 6972 6981 1 2 3 4 4 5 6 7 8
50 6990 6998 7007 7016 7024 7033 7042 7050 7059 7067 I 2 3 3 4 5 6 7 8
51 7076 7084 7093 7101 7110 7118 7126 7135 7143 7152 I 2 3 3 4 5 6 7 8
52 7160 7168 7177 7185 7193 7202 7210 7218 7226 7235 1 2 2 3 4 5 6 7 7
53 7243 7251 7259 7267 7275 7284 7292 7300 7308 7316 1 2 2 3 4 5 6 6 7
54 7324 7332 7340 7348 7356 7364 7372 7380 7388 7396 1 2 2 3 4 5 6 6 7
oI 1 2 I 3 141 6 I 6 7 I 8 1 9 11 213 41 6 16 71B 9

FOUR-FIGURE LOGARITHMS 823
I
- -MEAN DIFFERENCES
0 1 I2 3 4 5 6 7 8 9 11 2 13 41 5 16 7 81 9
55 7404 7412 7419 74:!7 7431i 7443 7451 7459 7466 4774 I 2 2 3 4 5 5 6 7
56 7482 7490 7497 7505 7513 7520 752!! 71i3fi 7;i4~ 7551 1 ~ 2 3 4 5 5 6 7
57 7559 7.'\66 7574 7582 7!i1l9 7li97 7fi04 7612 7619 7627 I 2 2 :I 4 5 5 6 7
58 7634 7642 764!! 7657 7664 7672 7679 7U86 7694 7701 I 1 2 :I 4 4 5 6 7
59 7709 7716 7723 7731 7738 774-5 7752 77(j() 7767 7774 1 1 2 3 4 4 Ii (l 7
60 7782 7789 779U 7803 7810 7818 7X;?il 7tl:J2 78:10 7846 I I .,
:I 4 4 Ii 6 6
61 7853 7860 7868 7875 7882 7889 7896 700:) 7910 7917 1 I
.,
:1 4 4 5 6 6
62 7924 79:U 7938 7945 7952 7959 7966 7973 7980 7987 1 I :! :3 :1 4 n. f) 6
63 7993 8000 8007 S014 8021 8028 8035 8041 8U48 8051i 1 I 2 :3 3 4 Ii 5 Ii
64 S062 8069 8075 8082 8089 8096 8102 8109 8116 8122 1 1 :! :1 3 4 5 5 6
65 8129 8136 8142 8149 8156 8162 8169 8176 8182 8189 I I 4 2 3 3 5 Ii 6
66 8195 8202 8209 8215 8222 8228 8235 8241 8248 8254 1 11 4 .):1 3 5 5 6
67 8261 8267 8274 8280 8287 8293 8299 8306 8:112 83 H) 1 I
\2 4 3 3 5 5 6
68 8325 8331 8338 8344 8351 8357 8363 8370 8376 8382 1 I 2 3 3 4 4 Ii 6
69 8388 8395 8401 8407 8414 8420 8426 8432 8439 8445 1 1 2 2 3 4 4 .5 Ii
70 8451 8457 8463 8470 8476 8482 8488 8494 8500 2 2 3 8506 1 I 4 4 [) 6
71 8513 8519 8525 8531 8537 8543 8549 8651) 8561 8567 1 1 2 2 3 4 4 :; 5
72 8573 8579 8,'\8.5 8591 8597 8603 8609 861.5 8621 S6!!7 1 1 2 2 3 4 4 5 5
73 8633 8639 8645 8651 8657 8663 8669 8675 8681 8686 1 1 2 2 3 4 4 5 5
74 8692 8698 8704 8710 8716 8722 8727 8n3 8739 874-5 1 1 2 2 3 4 4 5 r;
75 8751 8756 8762 8768 8774 8779 8785 879118797 8802 1 1 2 2 3 3 4 5 5

76 8S08 8814 8820 8825 8831 8837 8842 88488854 88W 1 I :! 23 3 4 5 Ii
77 8865 8871 8876 8882 8887 8893 8890 8904 8910 8911j I I 2 2 :l 3 4
78
79
8921
8976
8927
8982
8932
8987
8938
8993
8943
8998
89i9
9004
8954
9009
8960 8965
9015 9020
8971
902'0
I 1 2 2 3
I 1 :! :) 3
3 4
3 4
" 4
4
5
5
Ii
80 9031 9036 9042 9047 9053 9058 9063 9069 9074 9079 I 1 2 2 3 3 4 4 Ii
81 9085 9090 9096 9101 9106 9112 9117 9122 9128 9133 1 I 2 2 3 3 4 4 (j
82 9138 9143 9149 9154 9159 9165 9170 9175 9180 9186 1 1 :! 2 3 3 4 4 5
83 9191 9196 9201 9206 9212 9217 9222 9227 9232 9238 1 I 2 2 3 3 4 4 5
84 9243 9248 9253 9258 9263 9269 9274 9279 9284 9289 1 I 2 2 3 3 4 4 5
85 9294 9299 9304 9309 9315 9320 9325 9330 9335 9340 1 1 3 3 4 4 !'i
86 9345 9350 9355 9360 9365 9370 9375 9380 9:~85 9390 I I 2\2
2 2 3 3 4- 4 5
81 9395 9400 9405 9410 9415 9420 9425 9430 9435 9440 0 1 1 2 2 3 :I 4 4
88 9445 9450 9455 9460 9465 9469 94;-4 9479 9484 9489 0 1 I 2 2 3 :J 4 4
89 9'l94 9499 9504 9509 9513 9518 9523 952H 9.133 (1)38 0 1 1 ,2 2 3 :l 4 4
I
90 9542 9547 9552 9557 9562 9566 9571 9576 9581 9586 0 1 I 2 ~~ 3 3 4 4
91 9590 9595 9600 9(;05 9609 9614 9619 9624 9628 9633 0 I I 2 2 3 3 4 4
92 9638 9643 9647 9652 9657 9661 9666 9671 9675 9680 0 1 1 2 2 3 3 4 4
93 9685 9689 9694 9699 9703 9708 9713 9717 9722 9727 0 I I 2 2 3 3 4 4
94 9731 9736 9741 9745 9750 9754 9759 9763 9768 9773 0 1 1 2 2 3 3 4 4
95 9777 9782 9786 9791 9795 9800 9805 9809 9814 9818 {) 1 I 2 2 3 3 4 4.
96 9823 9827 9832 9836 9841 9845 9850 9854 9859 9863 0 1 1 2 2 3 3 4 4
97 9868 9872 9877 9881 9886 9890 9894 9899 9903 9908 0 1 I 2 2 3 3 4. 4-
98 9912 9917 9921 9926 9930 9934 9939 9943 9948 9952 0 1 1 2 2 3 3 4 4.
99 9956 9961 9965 9969 9974 9978 9983 9987 9991 9996 0 I I 2 2 3 3 3 4
-1'0 11 I 2 314 j""51
~ I 7 1 8 9 2 31 4 1 5 1 6 '1 8 9 11
A, 17. FIVE-FIGURE LOGARITHMS*
'" The five-figure logarithm tables (but in a modified set-out) are taken from
E. Hope, The Chemists' Year Book, 1939. and are reproduced by kind permission
of the publishers, Messrs. Sherratt and Hughes, Timperley, Cheshire, England.
Permission to reproduce five-figure logarithm tables was also kindly granted
by Messrs. G. Bell and Sons Ltd., Portugal Street, London, W.C.2, England.
from their Synopsis of Applicable Mathematics by L. Silberstein, and also by
Dr. A. Lange from his Handbook of Chemistq, 2nd Edition, 1937 (Handbook
Publishers Inc .• Sandusky. Ohio. U.S.A.).
FIVE-FIGURE LOGARlTHMS: 100-150 R21i
PROPORTIONAL PARTS
I
1~~.____~0~~1____~2~3~~4_1~~5_:_~6~~7~8~~9~~1~2~3~41 5 6 7 8 9
100 00 O()O 043 087 130 173 217 26() 303 341i 380 4 S 13 17 21 2630 3i; :{O
101 432 475 518 561 604 647 n89 732 775 817 4 8 13 17 :l1 26 30 34 3D
102 860 903 945 988 030 072 115 157 199 242 4 8 13 17 21 25 30 34 38
103 01 284 326 368 410 452 494 536 578 620 662 4 8 13 17 21 25 30 34 38
104 703 745 787 828 870 912 953 995 036 078 4 8 12 17 21 25 29 34 38
105 02 119 160 202 243 284 .325 366 407 449 490 4 8 12 17 21 25 29 33 37
106 531 572 612 653 694 735 776 816 857 898 4 8 12 16 20 24 29 33 37
107 938 979 019 060 100 141 181 222 262 302 4 8 12 16 20 24 28 32 36
108 03 342 383 423 463 503 543 583 623 663 703 4 8 12 16 20 24- 28 32 36
109 743 782 822 862 902 941 981 021 060 100 4 8 12 16 20 24 28 32 36
110'04 139 179 218 258 297 336 376 415 454 493 4 8 12 16 20 24 27 31 35
111 532 571 610 650 689 727 766 805 844 883 4 8 12 16 20 28 27 31 35
112 922 961 999 038 077 115 154 192 231 269 4 8 12 16 20 23 27 31 3"
113 05 308 346 385 423 461 500 538 576 614 652 4 8 11 15 .19 23 27 30 34
114 690 729 767 805 843 881 918 956 994 032 4 8 11 15 19 23 26 30 34
110 06 070 108 145 183 221 258 296 333 371 408 4 8 11 15 19 28 26 30' 34
116 446 483 521 558 595 633 670 707 744 781 4 7 11 15 19 22 26 30 33
117 819 856 893 930 967 004 041 078 115 151 4 7 11 15 19 22 25 29 33
118 07 188 225 262 298 335 372 403 445 482 518 4 7 11 15 18 22 25 29 33
119 555 591 628 664 700 737 773 809 846 882 4 7 11 15 18 22 26 29 32
120 918 954 990 027 063 099 135 171 207 243 4 7 11 14 18 22 25 29 32
121 08 279 314 350 386 422 458 493 529 565 600 4 7 11 14 18 21 25 29 33
122 636 672 707 743 778 814 .849 884 920 955 4 7 11 14 18 21 25 28 32
123 991 025 061 096 132 167 202 237 272 307 4 7 11 14 18 21 25 28 32
124 09 342 377 412 447 482 517 552 587 621 656 3 7 10 14 18 21 24 28 31
125 691 726 760 795 830 864 899 934 968 003 3 7 10 14 17 21 24 28 31
126 10 037 072 106 140 175 209 243 278 312 346' 3 7 10 14 17 21 24 27 31
127 380 415 449 483 517 551 585 619 653 687 3 7 10 14 17 20 24 27 31
128 721 755 789 823 856 890 924 958 992 025 3 7 10 14 17 20 24 27 30
129 11 059 093 126 160 193 227 260 294 327 361 3 7 10 13 17 20 24 27 30
130 394 428 461 494 528 561 594 628 661 694 3 7 10 13 17 20 23 27 30
131 727 760 793 826 860 893 926 959 992 024 3 7 10 13 17 20 23' 26 30
132 12 057 090 123 156 189 222 254 287 320 352 3 7 10 13 16 20 23 26 29
133 385 418 450 483 516 548 581 613 646 678 3 6 10 13 16 20 23 26 29
134 710 743 775 808 840 872 905 937 969 001 3 6 10 13 16 19 23 26 29
135 13 033 066 098 130 162 194 226 258 290 322 3 6 10 13 16 19 22 26 29
136 354 386 418 450 481 513 545 577 609 640 3 6 10 13 16 19 22 25 29
137 672 704 735 767 799 830 862 893 925 956 3 6 10 13 16 19 22 25 28
138 988 019 051 082 114 145 176 208 239 270 3 6 9 13 16 19 22 25 28
139 14 301 333 364 395 426 457 489 520 551 1182 3 6 9 12 16 19 22 25 28
140 613 644 675 706 737 768 799 829 860 891 3 6 9 12 15 19 22.25 28
141 922 953 983 014 045· 076 106 137 168 198 3 6 9 12 15 18 21 25 28
142 15 229 259 290 320 351 381 412 442 473 503 3 6 9 12 15 18 21 24 27
143 534 564 594 625 655 685 715 746 776 806 3 6 9 12 15 18 21 24 27
144 836 866 897 927 957 987 017 047 077 107 3 6 9 12 15 18 21 24 27
145" 16 137 167 197 227 256 286 316 346 376 406 3 6 9 12 15 18 21 24 27
146 435 465 495 524 554 584 613 643 673 702 3 6 9 12 15 18 21 24 27
147 732 761 791 820 850 879 900 938 967 997 3 6 9 12 15 18 21 24 26
148 17 026 056 085 114 143 173 202 231 260 289 3 6 9 12 15 18 20 23 26
149 319 348 377 406 435 464 493 522 551 580 3 6 9 12 15 17 20 23 26
150 .609 638 667 696 725 754 782 '811 840 869 3 6 9 12 14 17 20 23 26
o 1 2 345 6789..1234516789
·00 000-·17 869
826 FIVE-FIGURE LOGARITHMS: 150-200
PROPORTIONAL PARTS
I
1-:-:::::-+",,:,,=,~0:-::1~~2:-:,3~:.;:4~1-:~5~+,:,:6~~7~~8~9~-:1;.....:;:2~3-:-=:4-1 5 6 7 8 ~
150 17 liM) 1\38 667 Iml\ 7~1i 71H 78~ 811 840 869 :I Ii 9 l!l 14 17 ~o :!3 2$
151 898 921; 955 984 013 041 070 099 127 156 :\ 6 9 11 14 17 20 23 26
152 18 184 213 241 270 298 327 355 384 412 441 3 6 9 11 14 17 20 23 25
153 469 498 526 554 583 611 639 667 696 724 3 6 9 11 14 17 20 23 25
154 752 780 80S S37 S05 S93 92l 9<i9 977 (105 3 G S II a 17 20 23 2$
155 19 033 061 089 117 145 173 201 229 257 28& 3 6 8 11 14 17 20 22 2$
156 312 340 368 396 424 451 479 507 535 562 3 6 8 11 14 17 20 22 ~~
157 590 618 645 673 700 728 756 783 811 838 3 5 8 11 14 17 19 22 26
158 866 893 921 948 976 003 030 058 ,085 112 3 5 8 11 14 16 19 22 2t
159 20 140 167 194 222 249 276 303 330 358 385 3 5 8 11 14 16 19' 22 25
160 412 439 466 493 520 547 575 602 629 656 3 5 8 II 14 16 19 22 2i
161 683 710 736 763 790 817 844 871 898 925 3 5 8 11 13 16 19 22 24
162 951 978 005 032 059 085 112 139 165 192 3 5 8 11 13 16 19 21 24
163 21 219 245 272 299 325 352 378 405 431 457 3 5 8 11 13 16 19 21 ~t
164 484 511 537 564 5!!Q 6'17. 643 669 696 722 3 5 8 11 13 16 18 21 2
165 748 775 801 827 854 880 906 932 958 9sa 3 5 8 10 13 16 18 21 2~
166 22 011 037 063 089 115 141 167 194 220 246 3 5 R 10 13 16 18 21 23
167 272 298 324 350 376 401 427 453 479 505 3 5 8 10 13 16 18 21 23:
168 531 557 583 608 634 660 686 712 737 763 3 5 8 10 13 Hi 18 21 23 i
169 789 814 840 866 891 917 943 968 994 019 3 5 8 10 13 15 18 21 23
170 23 045 070 096 121 147 172 198 223 249 274 3 5 8 10 13 111 18 20 23
171 300 325 350 376 401 426 452 477 502 528 3 " 8 10 13 15 18 20 23
172 553 578 603 629 G54 679 704 729 754 779 3 " 8 10 13 15 18 20 23
173 80fi 830 855 880 905 930 955 980 005 030 2 5 7 10 13 15 18 20 23
174 24 055 080 105 130 155' 180 204 229 254 279 :) 5 7 10 12 15 17 20 22
175 304 329 353 :178 403 428 452 477 502 527 2 5 7 10 12 15 17 20 22
176 fifi1 mtl 601 62fi 1\50 674 699 724 748 773 2 fi 7, 10 12 15 17 20 2'2
177 797 822 846 S71 89:. 920 944 9(\9 993 018 2 li 7 10 12 15 17 20 22
178" 25 042 066 091 lUi 139 164 188 212 237 261 2 5 7 10 12 Iii 17 19 2.2
179 285 310 :134 35!; 3'82 406 431 455 479 503 2 [I 7 10 12 15 17 19 22
ISO ,,27 551 575 600 624 648 672 696 720 744 2 5 7 10 12 14 17 19 22
181 7GS 792 tlHl 840 Sfi4 888 912 935 959 983 :l Ii 7 10 12 14 17 19 22
182 26 007 031 055 079 102 126 lr.o 174 198 221 !l 5 7 10 12 14 17 19 21
,183 245 269 293 316 34(,) '364 387 411 435 458 2 5 7 10 12 14 17 19 21
184 482 505 529 GG3 U76 600 623 647 670 694 ;) 5 7 9 12 14 17 19 21
185 717 741 764 788 811 834 sa8 881 901i 928 2 5 7 9 12 14 16 19 21
186 9ril 975 998 021 045 068 091 114 138 161 2 5 7 9 12 14 16 19 21
187 27 184 207 231 254 '277 300 323 346 370 :{93 2 5 7 9 12 14 16 19 21
188 416 439 462 485 508 531 551 577 600 623 2 I) 7 9 12 14 16 18 21
189 64G 6ti9 692 715 738 7QI 784 807 829 852 2 Ii 7 9 II 14 16 18 21
190 875 898 921 944 967 989 012 035 058 OSl 2 r. 7 9 11 14 16 18 21
191 28 103 126 149 171 194 217 240 262 285 307 2 r. 7 9 11 14 16 1820
192 330 353 375 398 420 443 466 488 510 533 2 [) 7 !J II 14 16 18 20
193 556 578 601 623 646 668 691 713 735 758 :l 4 7 9 11 13 III 18 20
194 780 803 825 847 870 892 914 937 959 981 2 4 7 9 11 13' 16 18 20
195 29 003 026 048 070 092 115 137 159 181
203 2 4 7 9 11 13 16 IS 20
196 226 248 270 292 314 336 358 380 403
425 2 4 7 IJ 11 13 15 18 20
197 447 469 491 513 535 557 579 601 623
645 2 4 7 9 11 13 15 IS 2q
198 GG7 688 710 732 754 776 798 820 842
8G3 2 4 7 9 11 13 15 IS 2Q
199 885 907 929 951 973 994 016 03$ '060 OSl 2 4 7 9 11 13 1617 2q
200 3U 103 125 14G 168 190 211 233 255 276 298 247911 13 15 17 2J
~~~~0~~1~2~~3~~4~1 5
6 7 8 9 1 2 3 4"5" 6 7 8 D
-17609-,30 298
FIVE.FIGURE LOGARITHMS: 200-250 827
[ o 1 2 3 4 5 67891234
PROPORTIONAL PARTS
1 56789
30 103 12Q 146 168 190 211 233 255 276 298 2 4 7 9 11 13 15 17 20
320 341 363 384 406 428 449 471 492 514 .) 4 6 9 II 13 Iii 17 19
535 557 578 600 621 (i42 664 (HIli 707 728 2' 4 ti 9 11 1:3 1.5 17 19
750 771 792 814 835 856 877 899 920 942 2 4 6 9 11 13151719
963 984 006 027 048 069 091 112 133 154 2 4 6 8 11 13151719
205 31 175 197 218 239 260 281 302 323 345 366 2 4 (j 8 11 13 15 17 19
206 387 408 429 450 471 492 513 534 555 576 2 4 «; 8 11 13 15 17 19
'207 597 618 639 660 681 702 723 744 765 785 2 4 6 8 10 13 15 17 )9
~~
806 827 848 869 890 911 931 952 973 994 2 4 6 8 10 13 15 17 19
32 015 035 056 077 098 118 139 160 181 201 2 4 6 8 10 12 15 17 19
~~O 222 243 263 284 305 325 346 366 387 408 2 4 6 8
2 4 6 8
10
10
12 14 16 18
,..11 428 449 469 490 510 531 552 572 593 6\3 12 14 16 18
~12 634'654675695715 736 756 777 797 818 2 4 6 8 10 12 14 16 H
~~3
,..14
838 858 879 899 919
33 041 062 082 102 122
940
143
960
163
980
183
001 021
203 224
2 4.
2 4
6
6
8
8
10
10
12
12
14
14
16
16
H
H
.215 244 264 284 304 325 345 365 385 405 425 2 4 6 8 10 12 14 16 18
~
:216 445 465 486 506 526 546 566 586 606 626 2 4 6 8 10 12 14 16 18
17 646 666 686 706 726 746 766 786 806 826 2 4 6 8 10 12 14 16 18
18, 846 866 885 905 925 945 965 985 005 025 2 4 6 8 10 12 14 Hi 18
,219 34 044 064 084 104 124 143 163 183 203 223 2 4 6 8 10 12 14 16 18
220 242 262 282 301 321 341 361 381 400 420 ~ 4 6 8 10 12 14 16 18
:221 439 459 479 498 518 537 557 5'17 596 616 2 4 6 8 10 12 14 16 18
.~ 635 655 674 694 713 733 753 772 792 811 2 4 6 8 10 12 14 16 18
'223 830 850 869 889 908 928 947 966 986 005 2 4 6 8 10 12 14 16 18
[224 35 025 044 064 083 102 122 141 160 180 199 2 4 6 8 10 12 14 15 17
,.225 218 238 257 276 295 315 334 353 372 392 2 4 6 8 10 12 14 15 17
~26 411 430 449 468 488 507 526 545 564 583 2 4 6 8 10 12 13 15 17
-.227 603 622 641 660 679 698 717 736 755 774 2 4 6 8 10 11 13 15 17
.228 793 813 832 851 870 889 908 927 946 965 2 4 6 S 10 11 13 15 17
1.229 984 003 021 040 059 078 097 116 135 154 2 4 6 8 9 11 13 15 17
,230 36 173 192 211 229 248 267 286 305 324 342 2 4 6 8 9 _ 11 13 15 17
~231 3&1 380 31l1l 418 437 455 ,174 49:1 511 530 2 4 6 8 II 11 13 15 17
-232 549 568 586 605 624 642 661 680 698 717 2 4 6 7 9 II 13 15 17
'233 736 754 773 791 810 829 847 866 884 903 2 4 6 7 911131517
:234 922 940 959 977 996 014 033 051 070 098 2 4 6 7 9 11 13 15 17
,235 37 107 125 144 162 181 200 218 236 254 273 2 4 6 7 9 II 13 15 17
'236 291 310 328 346 365 383 401 420 438 457 2 4 6 7 9 11 13 15 17
'237 475 493 511 530 548 566 585 603 621 639 2 4 5 7 9 11 13 15 16
! 238 658 676 694 712 731 749 767 785 803 822 2 4 5 7 9 11 13 15 16
:239 840 858 876 894 IlI2 931 949 967 985 003 2 4 5 7 '9 11 13 14 16
240 38 021 039 057 075 093 112 130 148 166 184 2 4 5 7 9 11 13 14 16
'241 202 220 238 256 274 292 310 328 346 364 2 4 5 7 9 11 13 14 16
24~ 382 399 417 435 453 471 489 507 525 543 2 4 {) 7 9 11 13 14 Hi
243 561 578 596 614 632 650 668 686 703 721 2 4 5 7 9 11 12 14 16
244 739 757 775 792 810 828 846 863 881 899 2 4 5 7 9 11 12 14 16
245 917 934 91i2 970 987 005 023 041 058 076 2 4 5 7 9 11 12 14 16
246 39 094 IH 129 14-6 164 182 199217 235 252 2 4 5 7 9 11 12 14 16
247 270 287 305 322 340 358 375 393 410 428 2 4 5 7 9 11 12 14 16
248 445 463 480 498 515 533 550 568 585 602 2 3 5 7 9 10 12 14 16
249 .620 637 655 672 690 707 724 742 759 777 2 3.5 7 9 10 12 14 16
250 794 811 829 846 863 881 898 915 933 950 2 3 5 7 I) 10 12 14 16
01284567891234 66789
·30 103-·39 950
828 FIVE-FlGtJRE LOG.AR1TB1tIS: 250-300
PROPORTIONAL PARTS
0 1 2 3 4 I) 6 '1 8 9 1 2 3 4 Ii 6 7 8 9
250 39 79~ 811 829 846 863 881 898 915 933 950 2 3 5 7 9 10 12 1~ 16
251 967 985 002 019 037 054 071 088 106 123 2 3 5 7 9 10 12 14 16
252 40 140 157 174 192 209 226 243 261 278 295 2 3 5 7 () 10 12 14 15
253 312 329 346 364 381 398 415 432 449 466 2 3 5 7 9 10 12 14 15
254 483 500 518 535 552 569 586 603 620 637 2 3 5 7 9 10 12 14 15
255 654 671 688 705 722 739 756 773 790 807 2 3 5 7 8 10 12 14 1/\
256 824 841 858 875 892 909 926 943 959 976 2 3 5 7 8 10 12 14 15
257 993 010 027 044 061 D78 096 III 128 146 2 3 5 7 8 10 12 13 15
258 41 162 179 196 212 229 246 263 280 296 313- 2 3 5 7 8 10121315
259 330 347 363 380 397 414 430 447 46i 481 2 3 5 7 8 10 12 13 15
26D 497 514 531 547 564 581 597 614 631 647 2 3 5 7 8 10 12 13 15
261 664 680 697 714 731 747 764 781 797 814 2 3 5 7 8 10 12 13 15
262 830 847 863 880 896 913 929 946 963 979 2 3 5 7 8 10 12 13 15
263 996 012 029 045 062 078 096 111 127 143 2 3 5 7 8 10 12 13 15
264 42 160 177 193 209 226 243 259 275 292 308 2 3 5 7 8 10 11 13 15
266 325 341 357 374 390 406 423 439 455 472 2 3 5 7 8 10 II 13 16
266 48& 504 521 537 553 570 586 602 619 635 2 3 5 7 8 10 11 13 15
267 651 667 684 700 716 732 749 765 781 797 2 3 5 6 8 10 11 13 Hi
268 813 830 846 862 878 894 910 927 943 959 2 3 5 6 8 10 11 13 15
269 975 991 008 024 040 056 072 088 104 120 2 3 5 6 8 10 11 13 14
270 43 136 152 169 185 201 .217 233 249 265 281 2 3 5 6 8 10 11 13 i4
271 297 313 329 345 361 377 393 409 425 441 2 3 5 6 8 10 11 13 14
272 457 473 489 505 521 537 553 569 585 600 2 3 5 6 8 10 11 13 14-
273 616 632 648 664 680 696 712 727 743 759 2 3 5 6 8 10 11 13 14
274 775 791 807 823 838 854 870 886 902 917 2 3 5 6 8 10 11 13 14
275 933 949 965 981 996 012 028 044 059 075 2 3 5 6 8 9 11 13 14
276 44 091 107 122 138 154 170 .185 201 217 232 2 3 5 6 8 9 11 13 14
277 248 264 279 295 311 326 342 358 373 389 2 3 5 6 8 9 11 13 14
278 404 420 436 451 467 483 498 514 529 545 :) 3 5 6 8 9 II 12 a
279 560 576 592 607 623 638 654 669 685 700 2 3 5 6 8 9 11 12 14
280 716 731 747 762 778 793 809 824 840 S55 2 3 5 6 8 9 11 12 14
281 871 886902917 932 948 963 979 99-1 010 2 3 5 6 8 () 11 12 14
282 45 025 040 056 071 086 102 117 133 148 163 2 3 5 6 8 9 II 12 14
283 179 194 209 225 240 255 .271 286 301 317 2 3 5 6 8 9 II 12 14
284 332 347 362 378 393 408 423 439 454 469 2 3 5 6 ~ 9 11 12 I4
285 484 500 515 530 545 560 576 591 606 621 2 3 5 6 8 9 II 12 14
286 637 652 667 682 697 712 728 743 758 773 2 3 5 6 8 9 11 12 14
287 788 803 818 834 849 864 879 894 'lI()9 1)24 2 3 5 6 8 9 11 12 14
288 939 954 969 984 000 016 030 046 06D 075 2 3 5 6 8 9 11 12 14
289 46 090 105 120 135 150 165 180 195 210 225 1 3 4 6 8 9 10 12 1:1
290 240 255 270 285 300 315 330 344 359 374 1 3 4 6 7 9 10 12 1:1
291 389 404 419 434 449 464 479 494 509 523 1 3 4 6 7 9 10 12 \:1
292 538 553 568 583 598 613 627 642 657 672 1 3 4 6 7· 9 10 12 1:l
293 687 702 716 731 746 761 776 790 805 820 1 3 4 6 7 9 10 12 13
294 835 &i9 864 879 894 909 923 938 953 967 1 3 4 6 7 o 10 12 13
295 982 997 012 026 041 056 070 085 100 114 1 3 4 6 7 9 10 12 1:1
296 47 129 144 159 173 188 20.2 :l17 232 246 261 1 3 4 G 7 9 10 12 13
297 276 290 305 319 334 349 363 378 392 407 1 3 4 6 7 9 10 12 13
298 422 436 451 466 480 494 509 524 538· 553 I 3 4 {) 7 9 10 12 13
299 567 582 596 611 625 640 654 669 683 698 1 3 4 6 7 9 10 12 13
300 712 727 741 756 770 784 799 813 828 842 1 3 4 6 7 9 10 12 13
0 1 2 3 4 5 6 7 8 9 1 2 3 4 6 6 "l 8 ~
·39 794-·47 842

FIVE-FIGURE LOG.ARITHl'rI8: 300-350 829
PROPORTIONAL PARTS
301
0 1 2
300 47 712 727 741
857 871 885
756
900
8 4
770
914
-5
784
929
6
799
943
7
813
959
8
828
972
9
842
986
1 2 8 4
1
1
3 4 6
3 4 6
5
7
7
6 7 8 9
9
9
10
10
12
12
13
13
302 48 001 015 029 044 058 073 087 101 116 130 1 3 4 6 7 9 10 11 13
303 144 159 173 187 202 216 230 244 258 273 1 3 4 6 7 9 10 11 13
304 287 302 316 330 344 359 373 387 401 416 1 3 4 6 7 9 10 11 13
305 430 444 458 473 487 501 li15 li30 li44 558 1 3 4 (I 7 9 10 11 13
306 572 586 601 615 629 643 657 671 686 700 1 3 4 6 7 9 10 11 13
307 714 728 742 756 770 785 799 813 827 841 1 3 4 6 7 8 10 11 13
308 855 869 883 887 911 926 940 954 968 982 1 3 4 6 7 8 10 11 13
309 996 010 024 038 058 066 080 094 108 122 1 3 4 6 7 8 10 11 13
: 310 49 136 150 164 178 192 206 220 234 248 262 1 3 4 6 7 8 10 11 13
311 276 290 304 318 332 346 360 374 388 402 1 3 4 6 7 8 10 11 13
312 415 429 443 457 471 485 499 513 527 541 1 3 4 6 7 8 10 11 13
313 554 568 582,596 610 624 638 651 665 679 1 3 4 6 7 8 10 11 12
314 693 707 721 734 748 762 776 790 803 817 1 3 4 6 7 8 10 11 12
315 831 845 859 872 886 900 914 927 941 955 I 3 4 5 7 8 10 11 12
316 969 983 996 010 024 037 051 085 079 092 1 3 4 5 7 8 10 11 12
317 60 106 119 133 147 161 174 188 202 215 229 I 3 4 5 7 8 10 11 12
318 243 256 270 284 297 311 325 338 352 365 1 3 4 5 7 -8101112
319 379 393 407 420 433 447 461 474 488 501 1 3 4 5 7 8 10 11 12
320 515 529 542 556 569 583 596 610 624 637 1 3 4 Ii 7 8 9 II 12
321 650 664 678 691 705 718 732 745 759 772 1 3 4 5 7 8 9 11 '12
322 786 799 813 826 839 853 866 880 893 907 1 3 4 5 7 8 9 11 12
323 920 934 947 961 974 987 001 014 028 041 1 3 4 5 7 8 9 11 12
324 51 054 068 081 09li 108 121 135 148 161 175 1 3 4 5 7 8 9 11 12
325 188 202 215 228 242 255 268 282 295 308 I 3 4 5 7 8 !l 11 12
328 322 335 348 362 375 388 402 415 428 442 1 3 4 5 7 8 9 11 12
327 455 468 481 495 508 521 534 548 561 574 1 3 4 5 7 8 9 11 12
323 587 601 614 627 640 654 667 680 693 706 1 3 4 5 7 8 9 11 12
329 720 733 746 759 772 786 799 812 825 838 1 3 4 5 7 ·8 9 11 12
330 851 865 878 891 904 917 930 943 957 970 1 3 4 5 7 8 9 11 12
331 983 996 009 022 035 048 061 0711 088 101 1 3 4 5 7 8 9 10 12
332 52 114 127 140 153 166 179 192 205 218 231 1 3 4 5 7 8 9 10 12
333 244 257 270 284 297 310 323 336 349 362 I 3 4 5 7 8 9 10 12
334 375 388 401 414 427 440 453 466 479 492 1 3 4 Ii 6 8 9 10 12
335 50~ 517 530 543 556 569 5S2 59li 608 621 1 3 4 5 6 8 9 10 12
336 634 647 660 673 686 699 711 724 737 750 1 3 4 5 6 8 9 10 12
337 763 776 789 802 815 827 840 853 866 879 1 3 4 5 6 8 9 10 12
338 892 905 917 930 943 956 969 982 994 007 1 3 4 5 6 8 9 10 12
339 53 020 033 046 058 071 084 097 110 122 135 1 3 4 5 6 8 9 10 12
340 148 161 173 186 199 212 224 237 250 263 1 3 4 5 6 8 9 10 11
841 275 288 301 314 326 339 352 364 377 390 1 3 4 5 6 8 9 10 11
342 403 415 428 441 453 466\ 479 491 504 517 I 3 4 5 6 S 9 10 11
343 529 542 555 567 580 593 605 618 631 643 1 3 4 5 6 8 9 10 11
344 656 668 681 694 706 719 732 744 757 769 1 3 4 5 6 8 9 10 11
345 782 794 807 820 832 845 857 870 882 895 1 3 4 5 () 8 {) 10 11
346 908 920 933 945 958 970 983 995 008 020 1 3 4- 5 6 8 {) 10 11
347 54033045058·070083 095 108 120 133 145 1 2 4 5 6 8 9 10 11
348 158 170 183 195 208 220 233 245 258 270 1 2 4 5 6 7 9 10 11
349 283 295 308 320 332 345 357 369 382 394 1 2 4 5 6 7 {) 10 11
350 407 419 432 444 456 469 481 494 506 518 1 2 4 5 6 7 9 10 II
0 1 2 3 4 5 6 7 8 9 1 2 3 4 6 8 7 8 9
·41712-·54 618
R30 FIVE-FIGURE LOGARITHMS: 350-400
PROPORTIONAL PARTS
0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 I
360 54 407 419 432 444 456 469 481 49"* a06 ,,[8 1 2 4 5 6 7 9 10 1
361 531 543 555 568 580 59:1 605 In 7 630 642 1 2 4 5 6 7 9 10 1
352 654 667 679 691 704 7J6 728 741 753 765 1 2 4 5 6 7 9 10 L
363 777 790 802 814 827 839 851 864 876 888 1 2 4 5 6 7 9 10 II
3M 900 913 925 937 949 962 974 986 998 011 1 2 4 5 6 7 9 10 11
355 55 023 035 047 Of~ 072 084 09(\ 108 121 133 1 2 4 5 '6 7 9 10 11
356 145 157 169 182 194 206 218 230 242 255 1 2 4 5 6 7 9 10 11
357 267 279 291 303 315 328 340 352 364 376· 1 2 4 5 6 7 9 10 11
358 388 400 413 425 437 449 461 473 485 497 1 2 4 6 6 7 8 10 11
359 509 522 534 546 558 570 582 594 606 618 1 2 4 6 6 7 8 10 11
360 630 642 654 666 678 691 703 n5 727 739 1 2 4- 5 6 7 8 10 11
361 751 763 775 787 799 811 823 835 847 859 1 2 4 5 6 7 8 10 II
362 871 883 895 907 919 931 943 955 967 979 1 2 4- 5 6 7 8 10 II
863 991 003 015 027 038 050 062 074 086 098 1 2 4 5 6 7 8 10 11
384 56 110 122 134 146 158 170 182 194 2U5 217 1 2 4 5 6 7 8 10 11
365 229 241 253 265 277 289 301 312 324 336 1 2 4 5 6 7 8 10 11
366 348 360 372 384 396 407 419 431 443 455 1 2 4 5 6 7, 8 9 10
367 467 478 490 502 514 526 538 549 561 573 1 2 4 5 6 7 8 910
368 585 597 608 620 632 644 656 667 679 :i91 1 2 4 ii 0 7 8 9 10
369 703 714 726 738 750 761 773 785 797 808 1 2 4 5 6 7 8 9 10
370 820 832 844 855 S67 879 891 902 914 926 1 2 4 /) 6 7 8 910
371 937 949 961 972 984 996 008 019 081 043 1 2 4 5 6 7 8 910
372 57 054 066 078 089 101 113 124 136 148 1Il9 1 2 3 5 6 7 8 9 10
373 171 183 194 206 217 229 241 252 264 276 1 2 3 II 6 7 8 9 to
374 287 299 3lQ 322 334 345 357 368 380 392 1 2 3 5 6 7 8 9 10
375 40a 415 426 43ll 449 461 473 484 496 1107 1 2 3 (j 6 7 8 9 10
876 519 530 542 553 565 57(l 588 600 611 623 1 2 3 5 6 7 8 9 10
377 634 646 657 669 680 692 703 715 726 738 1 2 3 5 6 7 8 9 10
378 749 761 772 784 795 807 818 830 841 852 1 2 3 5 6 7 8 9H
379 864 875 887 898 910 921 933 944 955 967 1 2 3 5 6 7 8 9 10
380 978 990 001 013 024 085 047 058 070 081 1 2 3 5 6 7 8 9 10
881 58 092 104 115 127 138 149 161 172 184 195 1 2 3 5 6 7 8 9 10
882 206 218 229 240 252 263 274 286 297 309 1 2 3 5 6 7 8 9 10
883 320 331 343 354 355 377 388 399 410 422 1 2 3 5 6 7 8 9 10
384 433 444 456 467 478 490 5O1 512 524 535 1 2 3 5 6 7 8 9 10
385 545 557 569 580 591 602 614 625 636 647 1 2 3 4 6 7 8 910
886 659 670 681 692 704 715 726 737 749 760 1 2 3 4 6 7 8 9 10
387 771 782 794 805 816 827 838 85{) 861 872 1 2 3 4 6 7 8 9 10
388 883894906 917 928 939 950 961 973 984 1 2 3 4 6 7-8 910
389 995 006 017 028 040 051 062 078 084 095 1 2 3 4 6 7 8 9 10
390 L9 106 118 129 140 151 162 173 IS4 195207 1 2 3 4 6 7 8 9 10
391 218 229 240 251 262 273 284 295 306 318 1 2 3 4 6 7 8 9 10
392 329 340 351 362 373 384 395406417428 1 2 3 4 6 7 8 9 10
893 439 450 461 472 483 494 506 517 528 539 1 2 3 4 6 7 8 9 10
394 550 661 572 583 594 605 616 627 638 649 1 2 3 4 5 7 8 9 10
895 660 671 682 693 704 715 726 737 748 759 1 2 3 4 5 7 8 9 10
896 770 780 791 802 813 ,824 S35 S46 857 868 1 2 3 4 5 7 8 9 10
897 879 890 901 912 923 934 945 956 966 977 1 2 3 4 5 7 8 9 10
398 988 999 010 021 032· 043 054 065 076 0S6 1 2 3 4 5 7 8 9 10
899 60 097 108 119 130 141 152 163 173 184 195 1 2 3 4 5 7 8 9 10
400 206 217 228 239 249 260 271 282 293 304 1 2 3 4 5 7 8 9 10
0 1 2 3 4 5 6 7 8 D 1 2 3 4 5 6 "I 8 ~
·54 407-,60 304
FIVE-FIGURE LOGARITHMS: 400-450 831
. PROPORTIONAL PARTS
I
400
401
0
60 206
314
1
217
325
2
228
336
3
239
347
4
~49
:11,8
5
~6O
~6!l
6
:!71
:179
7
282
390
8
29a
401
9
a04
4012
1 2 3 4
I 2 :1 4
1 2 3 4
-i1
5
6 7 8 9
7 8 II 10
6 8 910
402 423 433 444 455 466 477 4B7 498 riOO ~'i20 I 2 a 4 5 (j R !J \(1
403 530 541 1;52 56:1 574 584 59;' 606 617 627 I 2 :1 4 5 (j 8 !l 10
404 638 6.49 660 670 681 692 703 713 724 73/; 1 2 3 4 5 6 8 9 10
405 745 756 767 778 788 799 810 820 831 842 1 2 3 4 [j 6 7 9 10
406 853 863 874 88r. 895 006 917 927 938 949 1 !! a 4 5 6 7 010
407 959 970 981 991 002 013 023 034 045 055 1 :J 3 4 5 6 7 9 10
408 61 066 077 087 098 109 119 130 140 151 162 1 2 :1 4 5 6 7 !l 10
409 172 183 194 204 215 225 236 247 257 268 1 2 3 4 5 6 7 8 10
410 278 289 299 310 321 331 342 352 363 374 1 !!3 4 5 6 7 8 10
411 384 395 405 416 426 437 448 458 469 479 1 2 3 4 5 6 7 8 10
412 490 500 511 521 532 542 553 563 574 584 I 2 3 4 5 G 7 8 9
413 595 606 616 627 637 648 658 669 679 690 1 2 3 4 5 Ii 7 H 9
414 700 711 721 732 742 752 763 773 784 794 1 2 3 4 5 6 7 6 9
415 805 815 826 836 847 857 868 878 888 899 1 2 3 4 5 G 7 8 9
416 909 920 930 941 951 961 972 982 993 003 1 2 :l 4 5 6 7 6 !)
417 62 014 024 034 045 055 065 076 086 007 107 1 2 :1 4 Ii (j 7 8 r
418 118 128 138 149 159 170 180 190 201 211 1 2 :I 4 [) 6 7 8 !)
419 221 232 242 252 263 273 284 294 304 315 1 2 3 4 Ii G 7 8 9
420 325 335 346 3~6 366 377 387 397 408 418 1 2 :I 4 5 Ii 7 8 !)
421 428 439 449 459 469 480 490 500 511 521 1 .,2 3 4 5 (j7 8 II
422 531 542 552 562 a72 583 593 603 613 624 1 '< :J 4 5 6 7 8 9
423 634 644 655 665 675 685 696 706 716 726 1 2 3 4 5 (; 7 H· n
424 737 747 757 767 7q8 788 798 808 818 829 1 l! 3 4 5 6 7 8 9
425 839 849 859 870 880 890 900 910 921 931 1 2 3 4 5 1\ 7 8 9
426 941 951 961 972 982 992 002 012 022 033 I 2 3 4 5 6 7 :; 9
427 63 043 053 063 073 083 094 104 114 124 134 1 2 :I 4 5 6 7 8 9
428 144 155165 175 185 195 205 215 225 236 1 2 3 4 5 6 7 8 9
429 246 256 266 276 286 296 306317 327 337 1 :! 3 4 5 6 7 8 9
430 347 357 367 377 387 397 407 417428438 1 2 3 4 5 6 7 8 n
481 448 458 468 478 488 498 508 518 528 538 J 2 3 4 5 6 7 S 1I
432 548 558 568 579 589 599 609 619 629 639 1 2 3 4 5 6 7 8 9
433 649 659 669 679 689 699 709 719 729 739 1 2 3 4 5 6 7 8 9
434 749759769779'789 799 809 819 829 839 I 2 3 4 5 6 7 8 9
435 849 859 869 879 889 899 909 919 929 939 I 2 3 4 5 6 7 8 !I
436 949 959 969 979 988 998 008 018 028 038 1 :l :1 4 5 6 7 8 !I
487 64 048 058 068 078 088 098 108 118 128 137 1 2 :1 4 5 6 7 8 11
438 147 157 167 177 187 197 207 217 227 237 I 2 3 4 5 (j 7 8 1)
439 246 256 266 276 286 296 306 316 326 335 1 2 3 4 ,) 6 7 8 9
440 345 355 365 375 385 395 404 414 424 434 I ;) 4
2 5 6 7 8 9
441 444 454 464 473 483 493 503 513 523 532 1 2 3 4 5 6 7 8 !)
.442 542 552 562 572 582 591 601 611 621 631 1 2 3 5 6 7 8 !I
443 640 650 660 670 680 689 699 709 719 729 1 2 3 4'" 5 6 7 8 9
444 738·748758768777 787 797 807 816 826 1 2 3 4 5 6 7 8 9
445 836 846 856 865 875 885 895 904 914 924 1 2 3 4 5 6 7 8 9
446 933 943 953 963 972 982 992 002 011 021 1 2 3 4 5 6 7 S 9
447 65 031 040 050 060 070 079 089 099 108 lIS 1 2 3 4 5 6 7 8 9
448 128 137 147 157 167 176 186196205.215 1 2 3 4 5 6 7 8 9
44~ 225 234 244 254 263 273 283 292 302 312 1 2 3 4 Ii I) 7 8 9
450 321 331 341 350 360 369 379 388 398 408 1 2 3 4' 51 6 7 8 9
0 1 2 3 4 5 6 7 8 9 1 2 3 4 71 6 7 8 ~
·60 206-·65 408
PROPORTIONAL PARTS
0 1 2 3 4 5 6 7 8 9 1 2 8 4 5 6 7 8 9
450 65 321 331 341 350 360 369 379 388 398 408 1 2 3 4 5 6 7 8 9
451 418 428 437 447 456 466 475 485 495 504 1 2 3 4 5 6 7 8 0
452 514 523 533 543 ~52 562 57l 58l 591 600 1 2 3 4 5 6 7 8 0
458 610 619 629 639 648 657 667 676 686 696 1 2 3 4 1\ I) 7 8 9
454 706 715 725 734 744 753 763 772 782 792 1 2 3 4 5 6 7 8 0
455 801 8ll 820 830 839 849 858 868 877 887 1 2 3 ,t 1\ fj 7 8 9
456 896 906 916 925 03J 944 954 963 973 982 1 2 :1 4 5 Ii 7 8 0
457 992 001 011 020 030 039 049 058 06S 077 1 2 :i 4 5 f) 7 8 9
458 60 087 096 106 115 124 134 144 153 162 172 1 2 3 4 5 Ii 7 8 I
4U9 181 191 200 210 219 229 238 247 257 266 1 2 3 4 5 6 7 8 9
460 276285'295305314 323 332 342 351 361 1 2 3 4 5 6 7 8 8

461 370 380 389 398 408 417 427 436 {4.')455 1 2 3 4 5 6 7 8 8
462 464 474 483 492 502 511 521 530 5i19 54!) 1 ~ :1 4 5 6 7 8 s
463 55B 567 577 586 596 605 614 624 633 642 1 2 :l 4 5 6 7 8 8
464 652 661 671 680 689 699 708 717 727 736 1 2 3 4 5 6 7 7 II
465 745 755 764 773 783 ' 792 801 811 820 829 1 2 4
3 Ii 6 7 7 8
466 839 848 857 867 876 B85 894 004 913 922 1 2 4
3 5· 6 7 7 8
467 932 941 950 960 969 978 987 996 006 015 1 2 :I 4 5 6 6 7 8
468 67 025 034 043 052 062 071 080 089 099 108 1 2 3 4 5 6 6' 7 8
469 117 127 136 145 154 164 173 182 191 200 1 2 3 4 5 6 6 7 8
470 210 2i9 228 237 247 256 265 274 284 293 1 2 3 4 5 6 6 7 8
471 302 311 321 330 339 348 357 367 376 385 1 2 3 4 5 (i 6 7 8
472 394 403 413 422 431 440 449 459 468 477 _1 2 3 4 5 6 6 7 8
473 486 495 504 514 523 532 '541 550560569 1 2 3 4 5 5 6 7 II
474 578 587 596 605 614 624 633 642 651 660 1 2 :I 4 5 5 6 7 8
475 669 678 688 697 706 711i 724 733 742 752 1 2 :I 4 5 5 6 7 8
476 761 770 ~79 788 797 806 815 1:>25 834 843 1 2 3 4 5 5 6 7 l\
477 852 861 870 879 888 897 906 915 925 934 1 2 3 4 5 5 6 7 8
478 943 952 961 970 979 988 997 006 015 024 I 2 3 4 5 5 6 7 8
479 68 034 043 052 061 070 079 088 097 106 115 1 2 3 4 5 5 6 7 8
480 124 133 142 151 160 169 178 187 196205 1 2 3 4 5 5 6 7 8
481 215 224 233 242 251 260 2,,9 278 287 296 1 2 3 4 5 5 6 7 8
482 305 314 323 332 341 350 359 368 377 386 1 2 3 4 4 5 6 7 8
483 395404413422'431 440 449 458 467 476 1 2 3 4 4 5 6 7 8
484 485 494 502 511 520 529' 538 547 556 565 1 2 3 4 4 5 6 7 B
6 7 ~
485 574 583 592 601 610 619 628637646.655 1 2 3 4 4 5 6 7
486 664 673 681 690 699 708 717 726 735 744 1 2 3 4 4 5
487 753 762 771 780 789 797 806 815 824 833 1 2 3 4 4 5 6 7 8
488 842 851 860 869 878 886 895 904 913 922 1 2 3 4 4 5 6 7 8
489 901 940 949 958 966 975 984 993 002 011 1 2 3 4 4 5 6 7 8
490 69 020 028 037 046 055 064 073 082 090 099 1 2 3 4 4 5 6 7 8
491 108 117 126 135 144 152 161 170 179 188 1 2 3 4 4 5 6 7 B
492 197 205 214 223 232 241 249 258 267 276 1 2 3 4 4 5 6 7 8
493 285 293 302 311 320 329 338 346 355 364 1 2 3 4 4 5 6 7 8
494 373 381 390 399 408 417 425 434 443 452 1 2 3 4 4 5 6 7 8
495 461 ,469 478 487 496 504 513 522 531 539 h 2 3 3 4 5 6 7 !
496 548 557 566 1574 583 592 601 609 618 627 1 2 3 3 4 5 Ii 7 !
497 636 644 653 662 671 679 688 697 705 714 1 2 3 3 4 5 6 7 !
498 723 732 740 749 758 767 771l 784 793 801 1 2 3 3 4 5 6 7 8
499 810 819 827 836 845 854 862 871 880 888 1 2 3 3 4 5 6 7 8
-500 897 906 914 923 932

0 1 2 3 4
940
5'"
949 958 966 975
6 7 8 9
1 2 3 3
1 2 3 4
4
5
5 6 7 8
6 7 8 9
·65 321-·69 975
FlVE-FIGURE LOGARITHM8: 500 -550 833
PROPORTIONAL PARTS
0 1 2 3 4 56 7 8 9 1 2 3 4 51 6 7 8 9
500 69 897 906 914 923 932 949 958 966 975 I 2 3 3
940 4 5 6 7 8
501 984 992 001 010 018 027 038 044 053 062 1 2 3 3 4 5 6 7 8
502 70 070 079 088 096 105 114 122 131 140 148 1 2 3 3 4 5 6 7 8
503 157 165 174 183 191 200 209217226234 1 2 3 3 4 5 6 7 ~
504 243 252 260 269 278 286 295 303 312 321 1 2 3 3 4 5 6 7 8
505 329 338 346 355 364 372 381 389 398 406 I 2 3 3 4 Ii (j 7 8
506 415 424 432 441 449 458 467 475 4S! 492 1 2 3 3 4 5 6 7 8
507 501 509 518 526 535 5<l4 552 561 5(19 578 1 2 3 3 4 5 6 7 8
50S 586 595 603 612 621 629 638 646 6M 663 1 2 3 3 4 5 6 7 8
509 672 680 689 697 706 714 723 731 740 748 1 2 3 3 4 5 I) 7 8
510 757 766 774 783 791 800 808 817 825 834 1 2 3 3 4 5 6 7 8
511 842 851 859 868 876 885 893 902 910 919 1 2 3 3 4 5 6 7 8
512 927 935 944 952 961 969 978 986 995 003 1 2 3 :l 4 5 6 7 8
513 71 012 020 029 037 046 054 062 071 0711 088 1 2 :1 3 4 u (i 7 8
514 096 105 113 122 130 139 147 155 Hi4 172 1 2 3 :l 4 5 {\ .7 8
515 181 189 198 206 214 223 231 240 248 257 1 2 3 3 4 u (l 7 8
516 265 273 282 290 299 307 31u 324 332 341 1 2 3 3 4 5 6 7 !j
517 349 357 366 374 383 391 399 408 4-16 425 1 2 ;1 3., 4 5 6 7 8
518 433 441 4(;0 458 4-66 475 483 492 500 508 1 2 3 ~ 4 5 6 7 8
519 517 525 533 542 550 559 567 575 584 592 1 2 2 3 4 5 6 7 8
520 600 609 617 625 634 642 6r;o 659 667 675 1 2 2 3 4 5 6 7 8
521 684 692 700 709 717 725 734 742 750 759 1 2 2 3 4 5 Ii 7 8
522 767 775 784 792 SOO 809 HI7 821i S:l4- 842 1 2 Q :1 4- 5 6 7 7
523 850 858 867 875 883 892 900 908 917 925 1 2 "2 3 4 Ii 6 7 7
524 933 941 950 958 966 97(; 983 991 999 008 I 2 !! 3 4 Ii Ii 7 'i
525 72 016 024 032 041 049 057 066 074 082 090 1 2 :I
2 4 (, (j 7 7
526 099 107 115 123 132 140 148 156 IOu 173 1 2 :1
2 4 Ii 6 7 7
527 181 189 198 206 214 222 230 239 247 ~55 I 2 2 :1 4 Ii 6 7 7
528 263 272 280 288 296 304 313 321 329 3:li I 2 2 3 4 5 6 7 'i
529 346 354 362 370 378 387 39ii 403 411 419 I 2 2 ;I 4- /) 0 7 7
530 428 436 444 452 460 469 477 4R5 493 :J()) 1 2 2 ;l 4 Ii 7 7
(l
509 518 526 534 542 550 558 567 57!.i 583 1 :! 2 :l 4 (i 7 7
531
532
533
591
673
599
681
607
689
GIG
697
G~4
705
6~2
71;\
MO G48 ui)G GGii
7:!2 730 738 746
I
I
:l
2
2
2
3
:\
4
4,
"
u 6 7 7
5 Ii 7 7
534 754 762 770 779 787 795 803 811 S19 8:li 1 :! 2 :l 4 5 G 6 7
535 835 843 852 860 S68 876 S84 S9~ 900 90S 1 2 2 :I 4 5 G G 7
536 916 925 933 941 949 957 965 9n 981 989 J 2 2 3 4 u G G 7
537 997 006 014 022 030 038 046 054 062 070 I :l 2 :l 4 5 6 (j 7
538 73 078 086 094 102 111 119 127 I:Jii 14:1 151 I :l 2 :1 4 5 () 6 7
539 159 167 175 183 191 199 ~07 215 2:l3 231 1 2 2 :3 4 5 Ii 0 'i
540 239 247 2ii5 264 272 280 288 296 304 312 1 2 2 3 4 5 6 6 7
541 320 328 336 344 352 300 368 376 3B~ 392 1 .,2 :! 3 4 5 6 (l 7
542 400408416424 432 440 448 456 404 472 1 2 :l 4 [j6 (l 7
543 480 488 496 504 512 520 528 536 5H 552 I 2 2 :1 4 5 n 0 7
544 560 568 576 584 592 600 608 616 624 632 I 2 2 :l 4 5 6 G 7
545 fHn 648 656 664 672 679 687 695 703 711 1 2 2 3 4 (; 6 6 7
546 719 727 735 743 751 759 767 775 783 791 1 2 2 3 4 5 6 (J 7
547 799 807 815 823 830 838 846 854 80:! 870 1 2 2 3 4 5 6 6 7
548 878 886 894 002 910 918 926 933 941 949 1 !! 2 3 4 5 6 13 7
549 957 965 973 981 989 997 005 013 020 02~ 1 2 2 3 4 5 6 6 7
550 74 036 044 052 060 068 076 084 092 099 107 1 2 2 3 4 5 6 6 7
0 1 2 3 4 5 6 7 8 9 11 2 3 41'S 6 7 8 9
EE ·69 897-·74 107
834
"N"'"'' " ..... _
FIVE-FIGURE LOGARlTBl'IIS: 550-600
,
. '1
PROPORTIONAL PARTS
0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
550 74 036 044 052 060 068 076 084 092 099 107 1 2 2 3 4 5 6 6 7
551 115 123 131 139 147 155 162 170 178 186 1 2 2 3 4 5 6 6 7
552 194 202 210 218 225 233 241 249 257 265 1 2 2 3 4 5 6 6
II 653 273 280 288 296 304 312 320 327 335 343 1 2 2 3 4 5 5 6 ~
554 351 359 367 374 382 390 398 406 414 421 1 2 2 3 4 5 5 6 7
555 429 437 445 453 461 468 476 484 492 500 1 2 2 3 4 5 5 6 7
556 507 515 523 531 539 547 554 562 570 578 1 2 2 3 4· 5 5 6 7
557 586 593 601 609 617 624 632 640 648 656 1 2 2 3 4 5 5 6 7
558 663 671 679 687 695 702 710 718 726 73a 1 2 2 3 4 5 5 6 7
559 741 749 757 764 772 780 ~88 796 80:1 811 1 2 2 3 4 5 5 6 1
560 819 827 834 842 850 858 865 873 881 889 1 2 2 3 4 5 5 6 7,
561 896 904 912 920 927 935 943 950 958 966 1 2 2 3 4 5 5 6 7
562 974 981 989 997 005 012 020 028 035 043 1 2 2 3 4 5 5 6 7
563 75 051 059 066 074 082 089 097 101) 1I3 120 1 2 2 3 4 Ii 5 6 7
564. 128 136 143 151 159 166 174 182 189 197 1 2 2 3 4 Ii 5 6 7
665 205 213 220 228 236 I 243 251 259 266 274 1 2 2 3 4. 5 5 6 7
666 282 289 297 305 312 320 328 335 343 351 1 2 2 3 4 5 5 6 7
667 358 366 374 381 389 397 404 412 420 427 1 2 2 3 4 Ii5 6 7
568 435 442 450 458 465 473 481 488 496 504 1 2 2 3 4 5 Ii 6 7
569 511 519 526 534 542 M9 557 565 572 5SO 1 2 2 3 4 5 5 6 7
670 587 595 603 610 618 626 633 641 648 656 1 2 2 3 4 5 5 6 7
671 664 671 679 686 694 702 709 717 724 732 1 2 2 .3 4 5 5 6 7
572 740 747 755 762 770 778 785 793 SOO S08 1 2 2 3 4 5 5 6 7
573 815 823 831 838 846 853 861 868 876 884 1 2 2 3 4 5 5 6 7
574 891 899 906 914 921 929 937 944 952 959 1 2 2 3 4 Ii 5 6 7
575 967 974 982 989 997 005 012 020 027 036 1 2 2 3 4 5 6 6 7
578 76 042 050 057 065 072 080 087 095 103 110 1 2 2 3 4 5 5 6 7
577 118 125 133 140 148 155 163 170 178 185 1 2 2 3 4 5 5 6 7
578 193 200 208 215 223 230 238 245 253 260 1 2 2 3 4 5 5 6 7.
579 268 275 283 290 298 305 313 320 328 335 1 2 2 3 4 5 5 II 7
580 '343 350 358 365 373 380 388 395 403 410 1 2 2 3 4 4 5 ·6 7
581 418 425 433 440 448 455 462 470 477 485 1 2 2 3 4 4 5 6 7
582 492 500 507 515 522 530 537 544 552 559 1 I 2 3 4 4 5 6 7
583 567 574 582 589 597 604 612 619 626 634 1 I 2 3 4 4 5 6 7
584 641 649 656 664 671 678 686 693 701 708 1 1 2 3 4 4 5 6 7
585 716 723 730 738 745 753 760 768 775 782 1 I 2 3 4 4 5 6 7
586 790 797 805 812 819 827 834 842 849 856 1 I 2 3 4 4 5 6 7
587 864 871 879 886 893 901 908 916 923 930 I 1 2 3 4 4 5 6 7
58S 938 945 952 960 967 975 982 989 997 004 1 1 2 3 4 4 5 6 7
589 77 012 019 026 034 041 048 056 063 070 078 I 1 2 3 4 4 5 6 7
690 085 093 100 107 115 122 129 137 144 151 1 1 2 3 4 4 5 6 7
591 159 166 173 lSI 188 195 203 210 218 225 1 1 2 3 4 4 5 6 7
592 232 240 247 254 262 269 276 283 291 298 1 1 2 3 4 4 5 6 7
593 305 313 320 327 3~5 342 349 357 364 371 1 1 ;! 3 4 4 5 6 7
594 379 386 393 401 408 415 422 430 437 444 1 1 2 3 4 4, 5 6 7
595 452 459 466 474 481 488 495 503 510 517 1 1 2 3 4 4 5 6 7
596 525 532 539 546 554 561 568 576 583 590 1 1 2 3 4 4 5 6 7
597 597 605 612 619 627 634 641 648 656 65:\ 1 1 2 3 4 4 5 6 7
598 670 677 685 692 699 706 714721728735' I 1 2 3 4 4 5 6 7
599 U:J 750 757 764 772 779 786 793 801 808 1 I 2 3 4 4 5 6 7
815 822 830 837 844 851 859 866 875 880 1 1 2 3 4 4 5 6 7
J!22..
0 1 2 3 4 s- 6 'i 8 9 1 2 3 4 5 6 7 8 9
·14 036-·71 880
FlVE-FlGURE LOGARl'rIntS: 600-650 835
PROPORTIONAL PARTS
0 1 2 3I5 I6 7 8 9
4 1 2 3 4 51 6 7 8 9
600 77 815 822 830 837 844I 851 850 866 945
887 895 902 909 916 924 931
87:1 880
9,,2 9~8
I 1 2
1 1 2
3
3
4 4
4
5
5
6
6
7
7
601
602
603
960 967 974 981 988
78 032 039 046 053 061
99(;
068
003 010 017 024
075 082 089 096
I 1 :!
1 1 2
3
3
.
4
4
5
4,
5
5 6
6 6
G
604 104 III 118 125 132 140 147 154 161 168 1 1 2 3 4 4 5 6 6
605 176 183 190 197 204 211 219 2~6 23~ 240 1 1 2 3 4 4 5 Ii 6
606 247 254 262 269 276 283 290 297 305 31:! 1 1 :!3 4 4 5 6 6
607 319 326 333 340 347 355 362 3G9 376 383 1 1 2 3 4 4 5 6 6
608 390 398 405 412 419 426 433 440 447 455 1 1 2 3 4 4 5 6 6
609 462 469 476 483 490 497 504 512 519 526 1 1 2 3 4 4 5 6 6
610 533 540 547 554 561 569 576 583 590 597 1 1 2 3 4 4 5 6 6
611 604 611 618 625 633 640 647 654 661 668 1 1 2 3 4 4 5 6 Ii
812 675 682 689 696 704 711 718 725 732 739 1 1 2 3 4, 4 5 6 6
813 746 753 760 767 774 781 789 791l 803 810 1 I 2 :I 4 4 Ii 6 6
614 817 824 831 838 845 852 859 86(i 873 880 I 1 2 3 4 4 5 6 6
616 888 895 902 909 916 923 930 937 944 951 1 1 2 3 4 4 1\ (i6
816 958 965 972 979 986 993 000 001 014 021 1 1 :! 3 4 4 Ii G G
817 79 029 036 043 050 057 064 071 078 085 092 1 1 :l 3 4 4 5 6 6
818 099 106 113 120 127 134 141 148 15:' ]lj~ 1 I :l 3 4 4 5 (i 6
819 169 176 183 190 197 204 211 218 z;!5 232 1 1 :l 3 4 4 5 6 6
820 239 246 253 260 267 274 281 288 295 302 1 1 2 3 4 4 5 6 6
821 309 316 323 330 337 344 351 358 365 372 1 1 2 3 4, 4 5 (; 6
822 379 386 393 400 407 414 421 428 4-:1:; 44-2 1 1 2 3 4 4 5 G 6
823 449 456 463 470 477 484- 491 498 505 512 1 1 2 3 3 4 (j 6 6
624 518 525 532 539 546 553 560 567 574 581 1 1 2 3 3 4 5 G 6
626 588 595 602 609 616 623 630 637 6U 650 I 1 2 3 3 4 5 6 6
826 657 664 671 678 685 692 699 706 713 720 1 1 2 3 3 4 5 6 6
827 727 734 741 748 754 761 768 775 782 789 I I :! 3 3 ·1 Ii (l G
828 796 803 810 817 824 831 837 844 851 858 I I 2 :l 3 4 5 6 6
829 865 872 879 886 893 900 9{)6 913 920 927 I. 1 2 :> 3 4 5 6 6
630 934 941 948 955 962 969 975 982 989 996 1 1 2 3 3 4 5 6 6
631 80 003 010 017 024030 037 044 051 058 065 1 1 2 3 3 4 (; 6 6
632 072 079 085 092 099 106 113 120 127 134- 1 1 2 3 :I 4 5 5 6
633 140 147 154 161 168 175 182 188 195 202 1 1 2 3 3 4 5 5 6
634 209 216 223 229 236 243 250 257 264 271 1 1 2 3 3 4 5 5 6
635 277 284 291 298 305 312 318 325 332 33~ 1 1 2 3 3 4 5 5 6
686 346 353 359 366 373' 380 387 393 400 407 1 1 2 3 3 « 5 5 6
637 414 421 428 434 441 448 455 462 468 475 1 1 2 3 3 4 5 5 6
638 482 489 496 502 509 516 523 530 536 543 1 1 2 3 3 4 5 5 6
639 650 557 564 570 577 5S4 591 598 604 611 1 1 2 3 3 4 5 5 (i
640 618 625 632 638 645 652 659 66.5 672 679 1 1 2 3 3 4- 5 5 ~
641 686 693 699 706 713 720 726 733 740 747 1 1 2 a 3 4 5 Ii 6
842 754 760 767 774 781 787 794 801 808 814 1 1 2 3 3 4 5 5 6
643 821 828 835 841 848 855 862 868 875 88~ 1 1 2 3 3 4 5 5 6
644 889 895 902 909 916 922 929 936 942 949 1 1 2 3 3 4 I) 5 6
645 956 963 969 976 983 990 996 003 010 017 1 1 2 3 3 4 5 5 6
646 81 023 030 037 043 050 057 064 070 07i 084 I 1 2 3 3 4 5 (; 6
641 090 097 104 III 117 124 131 137 144 1:;1 I I 2 3 3 4 5
648 158 164 IiI 178 184 191 198 204 211 218 1 1 :! 3 3 4 5 55 ~
649 224 231 238 245 251 258 265 271 278 28;; 1 1 2 3 3 4 5 5 6
650 291 298 305 311 318 325 331 338 345 351 1 1 2 3 3 4 5 5 6
0 1 B 3 4 I)
I 6 7 8 9 1 2 3 4 Ii 6 1 8 8
·77 815-·81 351
836 FIVE-FIGURE LOGA1UTBMS: 650-700
I
PROPORTIONAL PARTS
0 1 2 3 4 5 6 7 8 9 1 2 S 4 I) 6 '1 8 ~
81 291 208305 311 318 3:!5' 331 338 345 351 1 1 4 5 5 {
650
651 358
425
305 371
431 438
378
445
385
451
:101 :i98
465
405
471
411
478
418
485
1 1
1 1
!l
2
3
3
2 3
3
3 ....
5 5 6
5
4 5 ~ ~
652 458 3
653 491 498 505 511 518 525 531 538 544 551 1 1 2 3 3
654 558 564 571 578 584 501 598. 604 611 618 1 1 2 3 3 4 5 5 ~
655 624 631 637 644 651 1157 . (\(;.l 671 ~77 684 1 1 2 3 3 4 5 5 6
656 690 697 704 710 717 723 730 737 743 750 1 1 2 3 3 4 5 5 e
657 757 763 770 776 783 790 796 803 809 816 1 1 2 3 3 4 5 5 e
658 823 829 836 842 849 856 862 869 875 882 1 1 2 3 3 4 5 5 6
659 889 895 902 908 915 921 928 935 941 948 1 1 2 3 5 5
660 954 961 90R 974 981 987 994 000 007 014 1 1 2 3
3
3
"
4 5 5 6
6
661 82 020 027 033 040 046 053 060 066 073 079 1 1 2 3 3 4 5 5 6
662 086092099 105 112 119 121) 132 138 145 1 1 2 3 3 4 5 5 6
663 Iii! 158 164 171 177 184 l!)l 197 2M 210 1 1 2 3 3 4 5 5 6
664 217 223 230 236 243 250 256 263 269 276 1 1 2 3 3 4 5 5 e
665 282 289 295 302 308 315 321 328 334 341 1 1 2 3 3 4 5 I) 6
666 347 354 360 367 314 380 387 393 400 406 1 1 2 3 3 4 5 5 6
667 413 419 426 432 4:19 445 452 458 465 471 1 1 2 3 3 4 5 5 6
668 478 484 491 497 ,,(:4 "'0 51 7 5~3 5:1O 536 1 1 2 3 3 4 I) 5 6
669 543 549 556 5u2 5u9 515 582 588 595 601 1 1 2 3 3 4 5 5 6
670 607 614 620 627 633 640 646 653 659 666 1 I 2 3 3 4 I) I) 6
671 672 679 685 692 698 705 711 718 724 730 1 1 2 3 3 4 5 5 6
672 737 743 750 756 763 7Ii!) 776 782 789 795 1 1 2 3 3 4 5 5 6
673 802 808 814 821 827 834 840 847 853 860 r 1 2 3 3 4 5 I) 6
674 866 872 879 885 892 898 905 911 918 924 1 1 2 3 3 4, 4, 5 6
675
676
930
995
937
001
943
008
950
014
95r.
020
963
027
91i9
033
975
040
982
046
988
052
1 1 2 3
I I 2 3
3
3
..4
4 5 6
4 5 6
677 83 059 065 072 078 085 091 097 104 110 117 1 1 2 3 3 4 4 5 6
678 123 129 136 142 149 155 161 168 174 181 1 1 2 3 3 4 4 I) 6
679 187 193 200 206 213 219 225 232 238 244 1 1 2 3 3 4 4 5 6
680
681
251
315
257
321
264
327
270
334
276
340
283
347
289 296
353_359
302 _308
366 372
1 1 2 3
1 1 2 3
3
3
4
4
4
4 ~ :
682
683
684
378
442
506
385
448
512
391
455
518
398
461
525
404
468
531
410
474
537
417 423
480 487
544 550
429 436
493 499
556 563
1 1 2 3
1 1 2 3
1 1 2 3
3
3
3
4
4
4
4
4 ~
4 5 6
:
569 575 582 588 594 601 607 613 620 626 I 1 2 3 3 4 4 Ii 6 •
685
686 632 G39 645 651 61)8 664 670 677 683 689 1 1 2 3 3 4 4 5 6
687 096 702 708 715 721 727 734 740 746 753 1 I 2 3 3 4 4 5 6
688 759 765 771 778 784 790 797 803 809 816 I 1 2 3 3 4 4 5 6
689 822 828 835 841 &17 853 860 866 872 879 1 1 2 3 3 4 4 5 e
1 4 4 5
4 4 5 ~
690 885 891 898 904 OlD 916 923 920 935 942 1 2 3 3
691 918 954 960 967 073 979 986 992 998 004 1 1 2 3 3
692 81 Oil 017 023 029 036 042 048 055 061 067 1 1 2 3 3 4 <1 5 6
693 073 080 086 092 098 105 111 117 123 130 1 1 2 2 3 4 4 Ii ~
136 142 148 155 161 167 173 180 186 192 1 1 2 2 3 4. 5 6
694
695 198 205 211 217 223 230 236 242 248 255 1 1 2 2 3
"4 4 5 6
696 261 261 273 280 286 292 298 305 311 317 1 1 2 2 3 ...
4 5 6
697 323 330 336 342 348 354 361 367 373 379 1 1 2 2 3 4 4. 5 6
698 386 392 398 404 410 417 423 429 435 442 1 1 2 2 3 4 4 5 6
699 448 454 460 466 473 479 485 491 497 504 1 1 2 2 3 4 4 5 6
700 510 516 522 528 535 541 547 553 559 566 1 1 2 2 3 4 4- I) 6
0 1 2 3 4 5 6 7 8 9 1 2 3 4/ 5 6 7 8 9
-81 291--84 566
FIVE-FiGURE LOG.ARITHMS: 700-750 837
PROPORTIONAL PARTS
5\
il
0 1 2 3 41-L\ 6 7 8 9 1 2 3 6 7 8 (I
700 84 510 516 522
701 572 578 584
528
590
5351541
597 603
54 7
609
553
615
559
621
566 1 1 2
628 1 1 2 ~I 4
4
4
4
5
5
6
6
702 634 64()' 646 652 65S 665 671 677 683 689 1 1 2 2 4 4 5 6
703
704
696 '702 70S 714
757 763 770 776
720 726
782 788
733
794
739
800
745
807
751 1 1 2 2
813 1 1 2 2 iI 4
4
4
4,
5
5
6
6
705 819 825 831 837 ~44 850 856 862 868 874 1 1 2 2 3 4 4 5 6
706 880 887 893 899 905 911 917924930.936 1 1 2 2 3 4 4" 5 6
707 942 948 954 960 967 973 979 985 991 997 1 1 2 2 3' 4 4 5 6
708 85 003 009 016 022 Q28 034 040 046 052 058 1 1 '2 2 3 4 4 5 6
709 065 071 077 083 089 095 101 107 114 120 1 1 2 2 3 4 4 5 6
710 126 132 138 144 150 156 163 169 175 181 1 1 2 2 4 4
oj, '4 ~ ~
3
711 187 193 199 205 211 217 224 230 236 242 1 1 2 2 3
712 248 254 260 266 272 278 285 291 297 303 1 1 2 2 3 4 4 5 f
713 309 315 321 327 333 339 345 352 358 364 1 1 2 2 3 4 4 5 5
714 370 376 382 388 394 400 ~06 412 418 425 1 1 2 2 3 4 4 5 1
715 431 437 443 449 455 461 467 473 479.485 1 1 2 2 3 4 4 5 5
716 491 497 503 510 516 522 528 534 540 546 1 1 2 2 3 4 4 5 5
717, 552 558 564 570 576 582 588 594 600 60s 1 1 2 2
.~ I
4 4 5 5
718 612 618 625 631 637 643 649 655 661 667 1 1 2 2 4 4 5 5
719 673 679 685 691 697 703 709715,721 727 1 1- 2 2 4 4 ? 5
720 733 739 745'~1 757 763 769 775 781 788 1 1 2 2 ' 3 4 4 5 5
721 794 800 806 812 818 824 830 836 842 848 1 1 2 2 3 4 4 5 5
722 854 86Q 866 872 878 884 890 896 902 908 1 1 2 2 3 4 4 5 5
723 914 920 926 932 938 944 950 956 962 968 1 1 2 2 3 4 4 5 5
724 974 980 986 992 998 004 010 016 022 028 1 1 2 '2 3 4 4 5 Ii
725 86 034 040 046 052 058 064 070 076 082 088 ,I 1 2 2 3 4 4
726 094 100 106 112 118 124 130 136 141 147 1 1 2 2 3 4 4 ~ ~
727 153 159 165 171 177 183 189 195 201 207 1 1 2 2 3 4 4 5 5
728 ,213 219 225 231 237 243 249 255 261 267 1 1 2 2 3 4 4 5 6
729 273 279 285 291 297 303 308 314 320 326 1 1 2 2 3 4 4 5 5
730 332 338 344 350 356 362 368 374 380 386 1 1 2 2 3 4 4 5 5
731 392 398 404 410 416 421 427 433 439 445 1 1 2 2 3 4 4 5 5
732 451 457 463 469 475 481 487 493 499 504 I 1 2 21 3 4 4 5 5
733 5H) 516 522 528 534 540 546 552 558 564 1 1 2 2 I 3 4 4 5 5
734 570 576 581 587 593 599 605 611 617 .623 1 1 2 21 3 4 4 5 5
735 629 635 641 646 652 658 664 670 676 682 1 1 2 2 3 4 4 5 I
736 688 694 700 705 711 717 723 729 735 741 I I 2 2 3 4 4 5 5
737 747 753 759 764 770 776 782 788 794 800 1 1 2 2 3 4 4 5 5
738 806 812 817 823 829 835 841 847 853 '859 1 1 2 2 3 4 4 5 5
739 864 870 876 882 888 894 900 906 911 917 1 1 2 2 3 4 4 5 5
740 923 929 935 941 947 953 958 964 970 976 1 1 2 2 3 4 4 5 5
741 982 988 994 999 005 011 017 023 029 035 1 1 2 2 3 4 4 5 5
742 87 040 046 052 058 064 070 075,.081 087 093 1 1 2 2 3 4 4 5 5
'743 099 105 III 116 122 128 134 140 146 151 1 1 2 2 3 4 4 5 5
744 157 163 169 175 181 186 192 198 204 210 1 1 2 2 3 4 4 5 5
745 216 221 227 233 239 245 251 256 262 268 1 1 2 2 3 3 4 Ii Ii
746 274 280 286 291 297 303 309 315 320 326 1 1 2 2 3 3 4- 5 5
747 • 332 338 344 350 355 '361 367·373379384 1 1 2 2 3 3 4 5 5
748 390 396 402 408 413 419 425 431 437 442 1 1'" 2 ,2 3 3 4 5 5
749 448 454 460 466 471 477 483 489 495 500 1 1 2 2 3 3 4 5 (t
750 506 512 518 523 529 535 541 547 552 558 1 1 2 2 3 3 4, 5 Ii
0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
EE* ·84 510- -87 558
&'18
. PROPORTIONAL PARTS
0 1 "2 S 4 I) 6 7 8 9 1 2 S 4 5 6 '1 8 I
760 87 606 512 51~ 1 2 2
~
523 529 535 541 547 552 558 1 3 3 4
'1111 564 570 576 581 587 593 599 604 610 616 1 1 2 2 3 3 4 g
'11i2 622 628 633 639 645 651 656 662 668 674 1 1 2 2 :I 3 4 Il e
'163 6SO 685 691 697 703 708 714 720 726 731 1 1 2 2 :I 3 4 5 e
764 787 743 749 754 760 766 772 777 783 789 1 1 2· 2 3 3 4 5 t
756 795 800 806 812 818 823 829 831l 841 846 l' 1 2 2 3 3 4 Ii (
'166
~ ~
852 858 864 869 875 881 887 892 898 904 1 1 2 2 3 3 4
'167 910 915 921 927 933 938 944 950 955 961 1 1 2 .2 3 3 4
768 967 973 978 984 990 996 001 007 018 018 1 1 2 2 3' S 4 5 Ii
769 ·88 024 030 036 041 047 053 058 064 07b 076 -I 1 2 2 3 3 4 5 Ii
780 082 087 093 098 104 110 116 121 127 133 1 1 2 2 3 3 4
5 ~
761 138 144 150 156 161 167 173 178 184 190 1 1 2 2 3 3 4
762 196 201 207 213 218 224 230 235 241 247 1 1 .2 2 3 3 4 : Ii
763 252 268 264 270 275 281 287 292 298 304 1 1 2 .2 3 3 4 Ii e,
764 309 315, 321 326 332 338 343 349 355 360 1 1 2 .2 3 3 4 5 e:
765 366 372
378 383 389 395 400 406 412 417 1 1 2 2 3 3 4 5 I
766 423 429
434 440 446 451 457 463 468 474 1 1 2 2 3 3 4 5 t
767 4SO 485
<WI 497 502 508 514 519 525 530 1 1 2 2 3 3 4
5 ~
768 1 2
~
6116 542
547 553 559 664 570 576 581 587 1 2 3 3 4
769 593 5p8 604 610 615 621 627 632 638 6t3 1 1 2 2 3 3 4 II
770
771
649 655 660 666 672 677 683' 689 694 700 1 1 2
1 1•2
2
2
3 3 t 5
3 4
!
772
778
705
762
818
711
767
824
717
773
829
722
779
835
728,
784
840
734
'190
846
739
795
852
745
801
857
750
807
863
756
812
868-
1 1 2
1 1 2
2
2
3
3
3
3 4 6
3 4 4 6
!
774 874 880 885 891 897 902 908 913 919 925 1 1 2 2 ~ 3 4 4 ~
'175
776
7'17
930
986
89 042
936
992
048
941
997
053
947
003
059
953 9118 964 969
009 014 000 025
064 070' 076 081
975 981
031 087
087 092
1 1 2
1 1 2
I 1 2
2
2
2
3
3
3
3
3
3
4
4
4 4
:~ ~
778
779
098
154
104
159
109
165
115
170
120 126 131 137
176 182 187 193
143 148
198 204
1 1 ~
1 1 2
2
lI'
3
3
3
3
4
4 :~
780 209 215 221 226 232 237 243 248 254 260 1 1 2 2 3 3 4 4 ~
781 265 271 276 282 287 293 2~8 304 310 315 1 1 2 2 3 3 4 '4
782 321 326 332 337 343 348 354 360 365 371 1 •1 2 2 3 3 4 4.. :
783 376 382 387 393 398 404 409 415 421 426 1 1 2 2 3 3 4 4
784 432 437 443 448 454 459 465 470 476 481 1 1 2 2 3 3 445
785
786
787
487
542
597
493
548
603
498
553
609
504
559
614
509
564
620
515
570
625
520
575
631
526 531
58t' 586
636 642
537
592-
647
1
1
1
I,
1
1
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4 4 6
:~
788 653 658 664 669 675 6SO 686 691 697 702 1 1 2' 2 3 3 4 4 6
789 708 713 719 724 730 735 741 746 752 757 1 1 2 2 3 3 4 4 5
1 1 2 2 3 4
790
791-
763
818
768
823
774
829
779
834
785
840
790
845
796
851
SOl
856
So7
862
812
867 1 1 2 2
3
3 S 4 !:
792
798
873
927
878
933
883
938
889
944
894
949
900
955
905
960
911
966
916
971
922
977
1
1
1
1
2
2
2
2
3
3
3 4
3 4
3 4, 4 6
!~
794 962 988 993 009 015 020 026 031
998 004 1 1 2 2 3
'195 90 037 0!!'2 «)48 053 059 064. 069 075 080 086 i 1 2 2 3 3 4 4' ~
796
7fr1
091 097 102 108 113 119 124 129 135 140 1 1 2 2
146 151 157 162 168 173 179 184·1S9 l,9S 1 1 2 2
3
3
3 4
3 4
,3 4
!~
798
! ~
200 206 211 ~7 222 227 233 238 ~44 249 1 1 2 2 3
'199 255 260 266 2 1 276 282 287 293 298 30!!' 1 1 2 2 3 3 4
800. 309 314 320 325 331 336 342 34'1 352 358 1 1 2 2 3 3 4 4 ~
0 1 2 3 4 5 6 7 8 9 1 2 3 4 Ii 6 7 8 9
·87 506-·90 368
FIVE-FIGURE LOOAlUTBMS.: 800-~ 839
PROPORTIONAL PARTS
800
0
9v 309
1
314
2
320 325
3 4
331
5
336
390
6 '1
342 347
396 401
8 9
352 358
407 412
1 2 3 4
1 1 2
1 I 2
2
2
5
3
6
3
3
'1
4
4
,
8 9
5
4 5
801 363 369 374 380 385 3
802 417 423 4~8 434 439 444 450 455 461 466 1. 1 22 3 3 4 4 5
803 472 477 4 2 488 493 499 504 509 515 520 1 1 2 2 3 3 4 4 5
8~ 526 531 536 5t2 547 553 558563 569'574 1 1 2 2 3' 3 4 4 5
805 t80 585 t90 596 601 607 612 617 623 628 1 1 2 2 3 3 4 4 5
806 634 639 644 650 655 660 666 671 677 682 1 1 2 2 3 3 4 4 5
807 687 693 698 704 709 714 720 725 730 736 1 1 2 2 3 3 4 4 5
808 741747'752757763 7,68 773 779 784 789 1 1 2 2 3 3 4 4 5
809 795 800 806 811 816 822 827 832 838 843 1 1 2 2 3 3 4 4 5
&10 849 854 859 865 870 875 881 886 891 897 1 1 2 2 3 3 4 4 5
811 902 907 913 918 924 929 934 940 945 950 1 1 2 2 3 3 4 4 5
812 956 961 966 972 977 982 988 993 998 004 1 1 2 2 3 3 4 4 5
813 91 009 014 020 025 030 036 041 046 052 057 1 1 2 2 3 3 4 4 5
814 062 068 073 078 084 089 094 100 105 no 1 1 2 2 3 3 4. 4 Ii
815
816
116
169
222
121
174
228
126
180
233
132
185
238
137
190
243
142
196
249
148
201
254
153 158
206212
259 265
164
217
270
1 1 2
1 1 2
1 1 2
2
2
2
3
3
3
3
4
4 ~.!
-817 3 3 4 4 Ii
818 275 281 286 291 297 302 307 312 318 323 1 1 2 2 3 3 4 4 Ii
819 328 334 339 344 350 355 360 365 371 376 1 1 2 2 3 3 4 4 li
820 381 387 392 39] 403 408 413 418 424 429 1 1 2 2 3 3 4 4 5
821 434 440 445 450 455 461 466 471 477 482 1 1 2 2 3 '3 4 4 5
822 487 492 498 503 508 1i14 519 524 529 535 1 1 2 2 3 3 4 4 5
823 540 545 551 556 661 566 572 577 582 587 1 1 2 2 3 3 4 4 5
824 693 598 603 609 614 619 624 630 635 640 1 1 2 2 3 3 4 4 5.
825 645 651 656 661 666 672 677 682 687 693 1 1 2 2
826 698
751
703
756
709
761
714
766
719
772
724
777
730
782
735
787
740
793
745
798'
1
1
1
1
2 2 . 33 3
3
~ 4
4 4
5
5
827 2 2 3 3 4 4 5
828 803 808 814 819 824 829 834 840 845 850 1 1 2 2 3 3 4 4 5
829 855 861 866 871 876 882 887 892 897 903 1 1 2 2 3 3 4 4 5
830 908 913 918 924 929 934 939 944 950 955 1 1 2 2 3 3 4 4. 5
831 960 965 971 976 98L 986 991 997 002 007 1 1 2 2 3 3 4 4 Ii -
832 92 012 018 023 028 033 038 044 049 054 059 1 1 2 2 3 3 4 4 5
833 065 070 075 080 085 091 096 101 106 III 1 1 2 2 3 3 4 4 5
834 117 122 127 1?2 137 143 148 153 158 163 1 1 2 2 3 3 4 4 5
835 169..l74 179 IS4 189 195 200 205 210 215 1 1 2 2 3 3 4 4 Ii
836 221 226 231 236 241 247 252 257 262 267 1 1 2 2 3 3 4 4 5
837 273 278 283 288 293 298 304 309 314 319 1 1 2 2 3 3 4 4 5
838 324 330 335 340 345 350 355 361 366 371 1 1 2 2 3 3 4 4 Ii
839
840
376 381
428433
387 392
438·443
397
449
402
454
407
459
412
464
418
469
423
474,
1 1 2 2
1 1 2 2
3
3
. 3 4 4 Ii
3 4 4 6
841 480 483 490 495 500 505 511 516 521 1i26 1 1 2 2 3 3 4 4 5
842 681 536 542 547 552 657 562 567 572 578 1 1 2 2 3 3 4 4 5
843 583 588 593 598 603 609 614 619 624 629 1 1 2 2 3 3 4 4 Ii
844 634 639 U45 650 655 660 665 670 675 681 1 1 2 2 3 3 4 4 5
845 68~ 691 696 701 706 711 716 722 '727 732 1 1 2 ·2 3 3 4 4 5
846 737 742 747 752 75S' 763 768 773 778 783 1 1 2 2 3 3 4 4 5
847 788 793 799 804 809 814 819 824 829 834 1 1 2 2 3 3 4 4 5
848 840 845 850 855 860 865 870 875 881 886 1 1 2 2 3 3 4 4 5
~9 891 896 901 906 911 916 921 927 932 937 1 1 2 2 3 3 4 4 5
850 942 947 952 957 962 967 973 978 9S3 988 1 1 2 2 3 3 4' 4 5
0 1 2 3 4 51 6 7 8 9 1 2 3 4 61 6 7 8 9
·90 309-·92 988
PROPORTIONAL PARTS
0 1 2 3 41_J 6 7 8 9 :J. 2 3 4 _!J 6 7 8 ~

850 92 942 947 952 957 962 967 . 973 978 983 988 1 1 2 2 3 3 4 4 f
851 993 998 003 008 013 018 024 029 034 039 1 1 2 2 3 3 4 4 t
85!! 93 044 049 054 059 064 069 075 080 085 090 1 1 2 2 3 3 4 4 5
853 095 100 105 110 115 120 125 131 136 141 1 1 2 2 3 3 4 4 5
854 146.151 156161166 171 176 181 186 192 0 1 2 2 3 3 4 4 Ii
855 197 202 207 212 217 222 227 232 237 242 0 1 2 2 3 3 4 4 Ii
856 247 252 258 263 268 273 278 283 288 293 0 1 2 2 3 3 4 4 f
857 298 303 3US 313 318 323 328 334 339 344 0 1 2 2 3 3 4 4 e
858 349354 359·364 369 374 379 384 389 394 0 1 2 2 3 3 4 4, 5
859 399 404 409 414 420 425 430 435 440 445 0 1 2 2 3 3 4 4 "5
860 450 455 ~60 ~65 470 ~75 4&0 485 ~90 495 0 1 2 2 3 3 4 4 5
861 500 505 510 515 520 526 531 536 541 546 0 1 2 2 3 3 4
862
863
551
601
556
606
561
611
5p6
616
571
621
576
626
581
631
586
636
591
641
596
646
0
0
1
1
2
2
2
2
3
3
3 4
3 4 4 5
!:
864 651 656 661 666 671 676 682 687 692 697 0 1 ·2 2 3 3 4 4 Ii
702 707 712 717 722 727 732 737 742 747 0 1 2 2 3 3 4
~
865
866
867
752 757 762
802 807 812
767
817
772
822
777
827
782
832
787
837
792
842
797 0 1 2 2
847 0 1 2 2
2
2
3 4
3 4 4 4
!
868 852 857 862 867 872 877 882 887 892 897 0 1 2 2 2 3 ~ 4, 4
869 902 907 912 917 922 927 932 937 942 947 0 1 2 2 2 3 3 4 4
870 952 957 962 967 972 977 982 987 992 997 0 1 1 2 2 3 3 4 4
871 94002 007 012 017 022 027 P32 037 042 047 0 1" 1 2 2 3 3. 4 4
872 052 057 062 067 072 077 082 086 091 096 0 1 1 2 2 3 3 4 4
873 101 106 III 116 121 126 131 136 141 146 0 I 1 2 2 3 3 4 4
874 151 156 161 166 171 176 181 186 191 196 0 1 1 2 2 3 3 4 4
,
875 201 206 211 216 221 226 231 236 240 245 0 1 1 2 2 3 3 4 4
876 250 255 260 265 270 275 280 285 290 295 0 1 1 2 2 3 3 4 4
877 300 305 310 315 320 325 330 335 340 344 0 1 1 2 2 3 3 4
878 3~9 354 359 364 369 374 379 384 389 394 0 1 1 2 2 \ 3 3 4
879 399 404 409 414 419 424 429 433 438 443 0 1 1 2 2 3 3 4 4
880 448 453 458 463 468- 473 478.48348849S- 0 1 1 2 2 3 3 4 4
881 498 503 507 512 517 522 527 532 537 542 0 1 1 2 2 3 3: 4 _4
882 547 552 557 562 567 571 576 581 586 591 0 1 1 2 2 3 3 4 4
883 596 601 606 611 616 621 626 630 635 640 0 1 1 2 2 3 3 4 4!
884
885
645
694
650
699
655
704
660
709
665
714
670 675 680 685 689
719 724 729 734
0 1
738 0 1 1 2
1 2 2
2.
3 3 4 4·
3 3 4 4,
.
886 743 748 753 758 763 76B 773 778 783 787 0 1 1 2 'Z 3 3 4 4'
887 792 797 &02 807 812 817 tl22 827 832 836 0 I" 1 2 2 3 3 4 4
888 841 846 851 856 861 866 871 876 880 885 0 1 1 2 2 3 3 4 4
889 890 895 900 905 910 915 919 924 929 934 0 1 1 2 2 3 4 4 4l I
890 93b 944 949 954 959 963 968 973 978 983 0 1 1 2 2 3 3 4 4•
'891 988 993 99B 002 007 012 017 022 027 032 0 1 J. 2 2 3 3 4 4
892 95 036 041 046 051 056 061 066 071 075 080 0 1 1 2 2 3 3 4 4
893 085 090 095 100 105 109 114 119 124 129 0 1 1 2 2 3 3 4 4~
894 134 139 143 148 153 158 163 168 173 177 0 1 1 2 2 3 3 4 4
895 182 187 19~ 197 202 207 211 216 221 226 0 1 1 2 2 3 3 4 4
896 231 236 240 245 250 255 260 265 270 274 '0 1 1 2 2 3 3 4 4
897 279 284 289 294 299 303 308 313 318 323 0 1 1 2 2 3 3 4 4
898 328 332 337 342 347 352 357 361 366 371 0 1 1 2 2 3 3 4 4
899 376 381 386 390 395 400 405 410 415 419 0 1 1 2 2 3 3 4 4
900 424 429 434 439 444 448 453 458 463 ·468 0 1 1 2 2 3 3 4 .4
0 1 2 3 4 6 6 7 8 9 1 2 3 4 6 6 7 8 9
·92 942- ·95 468
FIVE-FIGURE LOGARITHMS: 900-950 841
PROPORTIONAL PARTS
0 1 2 3 4 5 6 7 8 9 1 2 3 41 5 6 7 8 9
·900 95 424 429 434 439 444 448 453 458 463 468 0 1 1 2 2 3 3 4 4
901 472 477 482 487 492 497 501 506 511 516 0 1 1 2 2 3 3 4 4
902 521 525 530 535 540 545 550 554 559 564 0 1 I 2 2 3 3 4 4
903
904
569
617
574
622
578
626
583
631
588
636
593
641
598
646
602
650
607
655
612
660
0
0
1
1
1 2
I 2
2
2
3
3
3
3 ::
905 665 670 674 679 684 689 694 698 703 708 0 I I 2 2 3 3 4 4
906 713 718 722 727 732 737 742 746 751 756 0 I I 2 2 3 3 4 4
907 761 766 770 775 780 785 789 794 799 804 0 1 1 2 2 3 3 4 4
908 ,S09 813 818 823 828 832 837 842 847 852 0 1 1 2 2 3 3 4 4
,909 856 861 866 871 875 880 885 890 895 899 0 1 1 2 2 3 3 4 4
910 904 909 914 918 923 928 933 938 942 947 0 1 1 2 2 3 3 4 4
911 952 957 961 966 971 976 980 985 99Q 995 0 1 1 2 2 3 3 4 4
912 , 999 004 009 014 019 023 028 033 038 042 0 1 1 2 2 3 3 4 4
913
,914
96 047
095
05~ 057 061 066
099 104 109 114
071
H8
076 080 085 090 0
123 128 133 137 0
1
1
1
1
2
2 ;1 3
3
3
3
4
4
4
4
9~5
. 142 147 152 166 161 166 171 175 180 185 0 1 I 2 2 3 3 4 4
916 190 194 199 204 209 213 218 223 227 232 0 1 1 2 2 3 3 4 4
917 237 242 246 251 256 261 265 270 275 280 0 1 1 2 2 3 3 4 4
'918 284 289 291 298 303 308 313 3n 322 327 0 1 I 2 2 3 3 4 4
919 332 336 341 346 350 355 360 305 369 374 0 1 1 2 2 3 3 4 4
920 379 384 388 393 398 402 407 412 417 421 0 1 1 2 2 3 3 4 4
921 426 431 435 440 445 450 454 459 464 468 0 1 1 2 2 3 3 4 4
922 473 478 483 487 492 497 501 506 &11 515 0 I 1 2 2 3 3 4 4.
923 520 525 530 534 539 544 548 553 558 562 0 I 1 2 2 3 3 4 4-
924 567 572 577 581 586 591 595 600 605 609 0 I 1 2 2 3 3 4 4
925 614 619 624 628 633 638 642 647 652 656 0 1 1 2 2 3 3 4 4
926 661 666 670 675 680 685 689 694 699 703 0 t 1 2 2 P 3 4 4
927 708 713 717 722 727 731 736 741 745 750 0 1 1 2 2 3 3 4 4
928 755 759 764 769 774 778 783 78B 792 797 0 '1 1 2 2 3 3 4 4
929 802 806 811 816 820 825 830 834 839 844 0 1 1 2 2 3 3 4 4-
930 848 853 858 862 867 872 876 881 886 890 0 1 1 2 2 3 3 4 4
931 895 900 904 909 914 918 923 928 932 937 0 1 1 2 2 3 3 4 4
932 942 946 951 956 960 965 970 974 979 984 0 1 1 2 2 3 3 4 4
933 988 993 997 002 007 Oll 016 021 025 030 0 1 1 2 2 3 3 4 4
934 97 035" 039 044 049 053 058 063 067 072 077 0 1 1 2 2 3 3 4 4
935 081 986 090 095 100 104 109 Hoi 118 123 0 1 1 2 2 3 t 4 4
936 128 132 137 142 146 151 155 160 165 169 0 1 1 2 2 3 3 4 4
937 174 179 183 188 192 197 202 206 211 216 0 1 1 2 2 3 3
988
939
220
267
225
271
230
276
234
280
239
285
243
290
248
294
253
299
257
304
262
308
0
0
1. 1 2
1 1 2
2 I 3
2 3
3 !:
3 4 4
940 313 317 322 327 331 336 340 345 350 354 0 1 1 2 2 3' 3 4 4
941 359 364 368 373 377 382 387 391 396 400 0 1 1 2 2 3 3 4 4
1142 405 410 414 419 424 428 433 437 442 447 0 1 1 2 2 3 3 4 '4-
943 451 456 460 465 470 474 479 483 488 493 0 1 1 2 2 3 3
944 497 502 506 511 516 520 520 529 534 539 '0 1 1 2 2 3 3 !:
945 543 548 ~52 557 562 566 571 575 580 585 0 1 1 2 2 3 3 4 4
946 589 594 598 603 607' 612 617 621 626 630 0 1 1 2 2 3 3 4 4
947 635 640 644 649 653 658 663 667 672 676 0 1 1 2 2 3 3 4 4
948 681 685 690 695 699 704 708 713 717 722 0 1 I 2 2 3 3 4 4
949 727 731 736 740 745 150 754 759 763 768 0 1 .1 2 2 3 3 4 4
950 772 777 782 786 791 796 800 804 809 813 0 1 1 2 2 3 3 4 4
0 1 2 3 4 "6 6 7 8 9 1 2 3 41 6 6 7 8 9
·95 424-·97 813
842 FIVE-FIGURE LOG.ARlTBMS: 950-1000
PROPORTIONAL PARTS
0 1 2 3 4 5 I 6 7 8 9 1 2 3 4 5 6 7 8 !
950 772 777 782 786 791 795 800 804 809 813 0 1 1 2 2 3 3 4 4
951 818 823 827 832 836 841 845 840 855 859 0 1 1 2 2 3 3 4 4
952 864 868 873 877 882 886 891 896 900 905 0 1 1 2 2 3 3 4 4
953 909 914 918 923 928 932 937 941 946 950 0 1 1 2 2 3 3 4 4
954 955 959 964 968 '973 978 982 987 991 996 0 1 I 2 2 3 3 4 4
955 98 000 005 009 014 019 0231 028 032 037 041 0 1 1 2 2 3 3 4
956 046 050 055 059 064 068 073 078 082 087 0 1 I 2 2 g 3 4 4
957 ~ 091 096 100 105 109 114 1I~ 193 12.7 132 0 1 I 2 2 3 3 4, 4
958 137 141 146 150 155 159 164 168 173 177 0 1 1 2 2 3 3 4 ·4
959 182 186 191 195 200 204 209 214 218 223 0 1 I 2 2 3 3 4, 4
960 227 232 236 241 245 250 254 259 263 268 0 1 1 2 2 3 3 4 4
961 272 277 281 286 290 295 299 304 308 313 0 1 I 2 2 3 3 4 4
962 318 322 327 331 336 340 345 349 354 358 0 1 I 2 2 3 3 oj, 4
963 363'367 372 376 381 385 390 3M 399 403 0 1 1 2 2 3 3 4 4
964 408412417 421 426 430 435 439 444 448 0 1 1 2 2 3 3 4 4
965 453 457
462 466 471 I 475 480 484 '489 493 0 I 1 2 2 3 3 4 4
966 498 502
507 511 516 520 525 529 584 538 0 I I 2 2 8 3 4 4
967 543 547
552 556 561 565 570 574 579 583 0 1 1 2 2 3 3 4 4
968 688 592
597 601 605 610 614 619 623 628 0 1 1"'2 2 3 3 4 4
969 632 b37 641 646 65Q 655 659 664 668 673 0 1 1 2 2 3 3 4, 4
970 677 682 686 691 695 700 704 709 713 717 0 1 1 2 2 3 3 4 I
971 722 726 731 735 740 744 749 753 758 762 0 1 1 2 2 3 3' 4 4
'972 767 771 776 780 784 789· 793 798 802 807 0 1 1 2 2 3 3 4 4
973 811 816 820 825 829 834 838 843 847 851 0 1 1 2 2 3 3 4 4
974 856 860 865 869 874 878 883 [\87 892 896 0 1 1 2 2 3 3 4 4
975 900 905 909 914 918 923 927 932 936 941 0 1 1 2 2 3 3 4 4
976 Q45 949 954 958 963 967 972 976 981 985 0 1 "1 2 2 3 3 4 4
977 989 9M 998 003 007 012 016 021 025 029 0 1 1 2 2 3 3 4
978 99 034 038 043 047'052 056 061 065 069 074 0 1 I. 2 2 3 3 4 4
979 078 083 087 092 096 100 105 109 114 118 0 1 1 2 2 3 3 4 4
980 123 127 131 136 140 145 149 154 158 162 0 1 1 2 2 3 3 4 4
981 167 171 176 ISO 185 189 193 198 202 207 0 1 1 2 2 3 3 4 4
982 211 216 220 224 229 233 238 242 247 251 0 1 1 2 2 3 3 4, 4
983 255 260 264 269 273 277 282 286 291 295 0 1 1 2 "2 3 3 4 4
984 300 a04 308 313 317 322 326 330 335 339 0 1 1 2 2 3 3 4 4
985
986
987
" 344
388
432
348
392
436
352
396
441
357
401
446
361
405
449
366 370 374 379 383
410 414 419 423 427
454 458 463 467 471
0
0
0
1 1 2
1 1 2
1 1 2
.2
2
2
3 3
3 3
3 3
"44 4 4
4
988 476 480 484 489 493 498 502 506 511 515 0 1 1 2 2 3 3 4 4
989 520 524 528 533 537 542 546 550 555 559 0 1 1 2 2 3 3 4 4
990 564 568 572 577 581 585 590 594 599 603 0 1 l' 2 2 3 3 4,4
991 607 612 616 621 625 629 634 638 642 647 0 1 1 2 2 3 3 4,4
992 651 65~ 660 664 669 673 677 68~ 686 691 0 1 I 2 2 3 3 4 4-
993 695 699 704 708 712 717 721 726 730 734 0 1 1 2 2 3 3 3 4
994 739 743 747 752 756 760 765 769 774 778 0 1 1 2 2 3 3 3 4-
I
995 782 787 791 795 800 804 808·813 817 822 0 1 1 2, 2 3 3 3 4
996 826 830 835 839 843 848 852, 856 861 865 0 1 I 2 2 3 3 3 4
997 870 874 878 883 887 891 896 900 904 909 0 1 1 2 2 3 3 3 4
998 913 917 922 926 930 935 939 944 948 952 0 1 I 2 2 3 3 3 4
999 957 961 965 970 974 978 983 987 991 996 0 1 1 2 2 3 3 3 4
.
- 0 1 2 3 4 5 I 6 7 8 9 1 2 3 4 6 6 7 8 9
·97 'i72 - ,99 996
INDEX
The following abbrevi,ations are used : -
D. determination c colorimetric
v = volumetric e electrolytic
g == gravimetric gv gas-volumetric
Absorbents in gas analyais for: sodium thlosulphate (g), 533;
acetylene, 763; benzene, 763; by succinic acid (g), 534-5; in
carbon dioxide, 762; carbon felspar, 663. .
monoxide, 765; hydrogen, Ambler's combustion pipette, 769.
766; oxygen, 764; unsatur- Ambler's portable appantus, 755;
ated hydrocarbons, 762-3. use of, 755-9.
Absorptiometer, "Spekker " photo- Ammonia, D. 01: in ammonium
electric, 689; use of, 690. salts (v), 305; by Nessler's
Absorption spectra methods, 135; reagent (c), 699. .
apparatus for the D. of carbon Ammoniacal cuprous chloride re-
dioxide, 593-4. agent, -763, 765; nickel cy-
Accuracy, definition of, 185. anide reagent, 763.
Acetic acid,' determination of Ammonium, D. 01: as chloro-
strength of glacial (v), 297; platinate (gl, 574; by nitro-
densities of aqueous solutions meter or gas-volumeter (gv),
of, 788. 776.
Acidimetry and alkallmetcy, 44, 45; Ammonium molybdate reagent, 465;
theory of, 58. as catalyst, 409. ..
Acids: common, strengths of, 790 ; Ammotlium thiocyanate ablution:
densities at 20°C, 788; ionisa- preparation' of O.IN, 318.;
tion constants of, 7; ionisation standardisation of, 319.
of, 1; polybasic, 2, 9, 10; Ampere', 172.
preparation of standard, 278; Analysed samples, 792-8; British
pseudo, 60; strengths of, 6; -chemical standards, 7~3.; Brit-
strong, 2; weak, 2. ish standardised steel. 795;
Activity, 8; coefficient, 8; pro- Bureau of Standards, 795; for
duct, 14. students in metallurgical ana-
Adipic acid, 292. lysis (Ridsdale's), 792.
Adsorption, 147; indicators, appli- Analysis: elementary, samples and'
cations of, 312-17; indicators, solutions for, 791; schemes of
theory of, 9 3 . . study of, S16, S17.
Ageing, of precipitates, 148. Analytical chemistry, literature of,
Air baths, 231. 783-7.
Alizarin S, 541. Analytical reagents, 226, 792-8.
Alizarin yellow R, 68. Andrews' procedure, 438.
Alkali, standard, preparation of, Anhydrone, 593. '
286. Anodes, for electrolytic determina-
Alkaline pyrOgallol, 764. tions, 599-602.
Alwninon, 710. Anthranilic acid, 166.
Aluminium alloy. analysis of, 652. Antimony, D. of: as tetroxide (g),
.Almnini.um, D. of: as alumina (g), 518; as trisulphide (g), 517-8;
488; byaluminon (c), 710; by by chloramine:T (v), 460; by
oxine (g), 535; by oxine (v), iodine (v), 428: by potassium
4p3; by KI-KIO. (g), 533; by bromate (v), 452; by potas-
843
844 Ind~x
sium iodate (v), 442; by sensitivity, 197, 19B, 203 ;
potassium iodide - pyridine sensitivity curve of a, 203;
method (e), 721; by potassium theory oJ, 194.
permanganate (v), 432; by Balancing method, 676. ~
pyrogallol (g), 518-9; in anti- Barium, D. 01: (v), 332; as (" 0-
monates (v), 429. mate (g), 565-6; as s . hate
Antimony and tin in type metal, D. (g), 565-6.
of,432. ,Barium ,hydroxide, standard sohi-
Apparatus: Ambler's portable, tion of, 295.
755; bakelite, 222; Bone and Barium peroxide, analysis of, with
Newitt, 761 ; Bone ana potassium iodate (v), 445.
Wheeler, 759; catalogues of, Bases: densities of, at 20°C, 790;
787; fused silica, 222; glass, pseudo, 60; strengths of, 6;
221; heating, 230; in gas strong, 3; weak, 3.
analysis, 740; iron, 222 ; Basic ac:etate method. 157.
nickel, 222 ; Orsat-Lunge, Beakers, 221 ; .covers for, 236.,
751-2; platinum, 222, 274; Beer's law, 672; application of,
porcelain, 221; Shepherd,754; 673; deviations from, 674.
silver, 22~; U.S. Bureau of Benzidine, 575.
Mines, 761. n-B~nzildIDxime, 159.
Aqueous vapour, pressure of, 737. Benzoic acid, 292.
Arsenic, D. of: as ammonium a-Benzoinoxime (cupron), 162.
uranyl arsenate (g), 515, 517; Beryllium, D. 01: as oxide (g), 536 ;
as magnesium pyroarsenate (g); by ammonium nitrite (g), 537 ;
515-6; as MgNH,As0 4, 6H.O by tannin (g), 537-8; in admix-
(g), 517; as silver arsenate (v), ture with aluminium, 539.
326; as trisulpllide (g), 515-6 ; Bismuth, n: of: as oxyiodiPe (g),
by Gutzeit's method (c), 714- ·506-7 ; by cupferron (g), 507-
9; by potassium bromate (v), ~; by oxine. (g), 507-B;. by
451; by potassium iodate (v), potassium iodide (c), 727 ; .:by
442; by Sanger-Black-Gutzeit pyrogallol, (g), 506-7.
method (c), 719; in :1rsenates Bismuthate method, lor manganese,
(v),430. 35B.
Arsenious oxide, purification of, Blank determination, 1~0.
4}6; standard solution of, 416. Bleaching powder: D. of available
Arsenite and arsenate in admixtq.re, chlorine in (v), 426-7, (gv), 776.
D. of (g), 515. Boiling rods, 223.
Ascarlte, 592. Bone and Hewitt apparatus, 761.
Atomic weights, table of, 778. Bone and Wheeler apparatus, 759.
Available chlorine, 4:.l6. Borates, D. of, 688.
Azolitmin, 6B Borax: as standard substance, 284;
determination of (v), 303-4;
purification of, 2B4; titration
Back e.m.f., 174. with strong acid, 83.
Bakelite apparatus, 222._ Borda'S method of wllighing, 208.
Balance: air-damped, 217; arms, Boric acid, D. of: (v), 303-4.
inequality of, 209; care and Brass, analysis of, 625.
use of, 200; chainomatic, 216 ; Bromates, D. of: as silver bromide
construction of, 194; descrip; (g), 582; by sodium thiosul-
tion of analytical, 195; deter- phate (v) .. 423.
mination of sensitivity of, 203 ; Bromides, D. of: as silver bromide
magnetically damped, 217; (g), 576; by mercurous nitrate
maximum load of, 19B ; (v), 314; by mercurous per-
requirements of a good, 197; chlorate (v), 315.; by silver
Index 845
nitrate (v), 313; by Volhard's Calculations of gravimetric analysis
method (v), 322. - 19a,473. .
Bromo-cresol green, 6S. Calibration: of burettes, 250; of
Bromo-phenol blue, 6S, 316. flasks, 241; of gas burettes,
Bromo-thymol blue, 6S. 739; of pipettes, 244; of
Bronze, analysis of, 631. weights, 212.
Bronzes, composition of, 63l. Carbon dioxide in air, D. of, 298.
Buffer action, 38; effect and solu- Carbon in steel, D. of, 639.
bility, 42; solutions, 3S, 80S- Carbonate and bicarbonate, D. in
12. mixtures of (v), 301; and
Buffer solutions: Clark and Lubs, hydroxide, D. in mixtures of
S10; German and Vogel, Sll ; (v), 299.
McIlvaine, S09 ; Michaelis, Carbonates, D. of (g), 592; (v), 295.
810; Sorensen, .809; Soren- Carbosorb, 593.
sen-Walburn, 811; standards, Casserol~ 221.
S08; table of, SOS-12; Wal- CeDs: concentration, 101; oxida-
pole, SOS. tion-reduction, 102; voltaic,
.. Bumpin~ " of solutions, 223, 232. 101.
Bunsen valve, 351. Centrifuge, 227.
Bunte burette, 748; use of, 749. Cerie ammonium nitrate, 377; puri-
Buoyancy of air in weighing, 210. fication of, 378.
Burette: calibrator, 251; holders, Cerie ammonium sulphate, 377·8.
247; reader, 248., Cerie sulphate solutions :' oxidising
Burettes, 43, 246; Bunte, 74S: propeJ;ties of, 375; prepara-
calibration of, 250; Hempelt tion of O.lN, 378; standard-
745; Schellbach, 24S; isation of, by arsenious oxide,
standardised, 252; tolerances 380, '!?y Mrrous ammonium
of, 252; use of, 247-8; weight, sulphate, 381, by pure iron,
253 ; ~ith glass stop-cock, 380, by sodium oxalate, 382.
247; with pinch-cock, 247. Cerium, D. of: as ceric iodate (g),
Burners: Amal, 219; Bunsen, 558; by ferrous solution (v),
218; Fisher high temperature, 386; by potassium perman-
21S, 219; Meker, 21S-9 ; ganate (v), 386; by sodium
Pittsburgh universal, 21S; thiosulpbate (v), 386.
Tirrill, 218. Chemical factors, ~ 73; table of,
779-82.
Chloramine-T solutions: oxidising
CadmiUJll, D. of: as metal (e), 609, properties of, 459; prepara-
611; as molybdate (g), 50S-9; tion of O.IN, 459; standardi-
by t:l-naphthaquinoline (g), 50S- sation of, 459.
9; by oxine (g), 456; by Chlorates, D. of f, as silver chloride
oxine (v), 455; by potassiu~ (g), 5S0; by sodium thiosul-
iodate (v), 472; by pyridine phate (v), 423.
(g), 509-10; by pyridine- Chlorides. D. of: as silver chloride
ammonium thiocyanate (v), (g), 475-6; by mercurous
324; by quinaldinic acid (gr, perchlorate (v), 315; by silver
50S, 510. nitrate (v), 313; by Volhard's
CalcinQl, D. of.: as carbonate (g), method (v), 321.
490, 493; as oxide (g), 490, Chloroform, as indicator, 404, 438.
494; as oxalate monohydrate Chloro-phenol red, 68.
(g), 490, 564; by potassium Chloroplatinic acid reagent, 572.
permanganate (v), 352; in Chromate to chromic salt, 540.
felspat, 663; in limestone or Chromic sI&lt to chromate, 540.
dolomite, 657, 660. ChromiUJl)., D. of: as barium chro-
846 Index
mate (g), 540-1; as mercurous cyanate (~, 324; by quinal-
chromate (g), 539; as sesqui- dinic acid (g), 511, 513; by
oxide (g), 539; by diphenyl- salicylaldoxime (g), 511, 513;
carbazide (c), 709; by potas- by sodium diethyldithiocar-
sium cyanatll (g), 540; by bamate (e), 726-7; by sodium
potassium dichromate (v), 371 ; thiosulphate (v), 418, 419; in
by KI-1nOa {g), 539; in brass, 627; in German silver,
chromite, 371; In steel, 643-4. 634; in silver coinage alloy,
Chromous salts, reduction with, 3.99. 63S; in steel, 650, 652 .
. Cinchonine, 647. Copper-wckel alloy, analysis of, 616.
Cleaning of glass apparatus, 240. Copper oxide, combustion with, 710.
Coagulation values, 143. Coprecipitation, 147.
Cobalt, D. of: as metal (e), 547, 613; Coulomb, 172.
by l1-nitro-(3-naphthol (g), 547- Counter ions,143.
8; by a-nitroso-tl-I!-aphthol Cresol red, 68.
(c), 714; by a-D1~roso-~ Crucibles : Gooch, 263; ignition of,
naphthol, (g), 546-7; by oxin~ 265; Main Smith, 272; Mun-
(v), 456; by phenylthio- roe, 266; perforated screen
hydantoic acid (g), 547-8; by for, 277; platinum, 275: por-
pyridine (g), 547, 549; by celain, 221; porous porcelain
pyridine-ammonium thiocyan- filter, 267 ; quartz filter, 267 ;
ate (y), 323; in steel, 648 .• Rose,552; sintered' glass filter,
Colloidal slate, 140. 266; vitreosil Gooch, 266.
(lolo~eters : Duboscq, 864; Crushing of sample, 234.
hydrogen ion,' 896; plunger Cubic centimetre, 238.
type, 683. Cupferron, J 60.
Colorimetric analysis, 670; analysis, Cupron, 162.
theory of, 671 ; determinations, Cuprous splphate-,s-naphthol re-
general remarks upon, 697; D. agent, 766.
of elements and radicals, 730; Cuprous oxide, preparation of, 766.
titration, 682. Cyanides, D. of! as silver cyanide
Colour change interval, 58. (g), 578; by silver nitrate (v).
Colour measurement, classification 327-8. .
of methods of, 675.
Colour standards, permanent, 682, . Damping of balances, 217.
694-6. Decantation, washing by, 269.
Common ion effect, 11. Decomposition potential, 172.
Completeness of deposition, 177. Dehydrite, 593.
Complex formation reactions: dis- Deposition, completeness of, 177.
cussion of, 310; theory of, 89. Desiccators; 223, 225; use of, 224;
Complex ions, 22; instability con- vacuum. 224.
stants oi, 24. J)evarda's alloy, 306.
Conductivity water, 220. Dichlorofluotescein, 313.
Conductometric titrations, 132. Differential titration, 127.
Congo red, 68. Digestion, 148.
Control, running a, 190~ Di-iododimethylfluorescein, 316.
Copper, D. of! as cuprous thiocyan- Dilution lllethod, 675.
ate (g), 511-2; as 'metal (e), p-DimethYlruninO-benzal-rhodanine.
512, 606; by a-benzoinoxime 728.
(g), 511, 513; by ferrocyanide Dimethylglyoxime, 159.
(c), 726; by oxine (v) and (g), Diphenylamine, 114, 365; sodium
456; by potassium iodate (v), or barium sulphonate, 115, 365.
442; by pyridine (g), 511, 513 ; Diphenylbenzidine, 115, 366.
by 'pyridine-ammonium thio- DiPhenylcpbazide, 709.
Index 847
Diphenylthiocarbazone, 724-5. 'precipitation and complex-
Direct weighing, 201. formation reactions, 90.
Displacement titrations, 83. Eosin, 314. -
Dissociation constants: of acids, 5, Equilibrium constant. 4; of oxida-
7 ; of bases, 7. tion-reduction reactions, 104.
Distribution: coefficlent, 184 ; Equivalence point. 43, 58.
method, 182. Equivalent system. advantages of,
Dithizone, 724-5. " 53.
Dolomite, ahalysis of, 654. Equivalent weights. 45; of acids,
Draining time. 244. 45; of bases, 46; of oxidants,
Drying agents, 225; of precipitates, 51-2; of reductants, 51-2; of
270. salts, 46.
Duboscq colorimeter, 684; use of, Errors: absolute, 185; classifica-
685. " tion Of, 187; curve, 188;
Duplication method, 675, 682. determinate or constant, 187;
due to electrification, 209;
Electric ovens, 231. indeterminate or accidental,
Electrical units, 172. 187; in quantitative analysis,
Electrification due to wiping, 209. 185; in weighing 208; mini-
Electro-analysis, theory of, 171. misation of, 190; relative, 185.
Electrochemical equivalellt, 172; Ethyl ether. extraction with, 182;
series, 100. peroxides in, 605.
Electrode potentials, 98; potentials, Ethylenediamine, 167; use in
standard, 99, 100; reactions, separation of copper and mer-
175; vessels, 603. cury, 167. .
Electrodes: calomel, 119, 120, 131 ; Explosion and combustion, methods
cleaning of, 606; for electro- of aRalysis, 767.
lytic determinations, 598 ; Extemal indicators. 365.
glass, 125, 126, 131 ; hydrogen, Extraction: of solid mixtures by
118, 122; quinhydrone, 123; solvents, 181; of solutions by
Sand's, 600, 601; use and care solvents, 182.
of, 604. Evaporation, 236.
Electrolysis units, 596; rapid, 181, Evaporator. Moroney, 232.
608; slow, 181, 608.
Electrolytes, weak, 3. Faraday, 172; laws of, 171.
Electl'l:llytic apparatus, 596; de- Felspar. i-nalysis of, 661.
posit, character of, 596; deter- Ferric alum indicator. 319.
minations, 596; dissociation, Ferric iron, D. of: by sodium thio-
1 ; separation of metals, 179. sulphate (v), 436; by titanous
Electrolytic separations: copper salts (v), 392 (see under Iron).
and lead, 628'; copper and Ferric iron, reduction of: by
nickel, 615; copper and silver, hydrogen sulphide, 350; by
615. Jones' reductor, 346 ; by liquid
Electromotive force, 101; o:f voltaic amalgams, 394; by stannous
cell, 101. chloride, 345; by sulphurous
Electron diffraction methods, 135. acicf, 349; by zinc and sul-
Electropode, 135. phuric acid, 350.
Elementary analysis, samples and Ferricyanides, D. of: by sodium
solutions for, 791; scheme of, thiosulphate (v), 435.
816. Ferrocyanidcs. D. 01: by ceric
"Emulsoids, 142. sulphate (v), 383; . by chlora-
End points, 43; in neutralisation mine-T (v), 462.
reactions, 84; in oxidati~n Ferrous ammonium sulphate(Mohr's
reduction reactions, 112; in salt), purification of, 338
848 Index
Ferrous iron, D. of: by eerie Gas volumeter, Lunge's, 737.
sulphate (v), 380-1, 383; by Gas-volumetric analysis, 732, 771-7.
'manganic sulphat~ (v), 388; Gauss's method of double weighing,
by potassium dichromate (v), 208.
3G8; by potassium perman- Gels, 142. •
ganate (v), 344 (see under German silver, analysis of, 634.
Iron). Glassware: analytical, 221; clean-
Ferrous and ferric iron in are, D. of, ing of, 240.
370. Gold, D. of: as metal (g), 531-3.
Filter papers, 259; macerated, 262 ; Gooch crncjbles, 263; asbestos for,
incineration of, 272-3; table 263; heating of, 265; prepara-
, of Whatman's, 260. tion of, 263; vitreosiI, 266.
Filtration. 259; accelerated, 262; Gradation method, 676, 687.
with filter crucibles, 263 et seq. ; Graduated cylinders, 254.
w1th paper, 260-2; technique . Gram, 199.
of,269. • Gravimetric analysis: calculations
Flasks, conical (Erlenmeyer), 221. of, 473; forms in which ele-
Flocculation values, 143, 144. ments and radicals are com-
Fluorescein, 313. monly precipitated and weighed
Fonnates, D. of: by potassium as, 137; systematic, geneJ;'al
permangana~e (v), 361-2. , discu'ssion of, 500; technique
Fluorides, D. of: as calcium fluoride of, 258; theory of, 136.
(g), 5'78-9; as lead chloro- Gravimetric separations, simple:
fluoride (g), 578; as lead aluminium and chromium, 620;
chlorofluoride (v), 324; as calcium and barium, 622;
triphenyltin fluoride (g), 578-9. calcium and magnesium, 621.
Fluosilicates, D. of: (g), ..591. 657, 660; calcium and stron-
Fluxes, 237. tium, 623; calcium, strontium
Fonnates, D. of·:. by potassium and barium, 623; iron and
pennanganate (v), 361. aluminium, 617-9; iron and
Fractional combustion: of gases, chromium, 619 { manganese
770; pipette, 753. and zinc, 620 ; rticl<el and zinc,
Funnels, 22; Buchner, 228; Jena 621; sodium and potassium.
"slit sieve," 228; sintered 624, 664.
glass, 228-9. Greases for glass stop-cocks, 249.
a~Furi1dio:xime, 159. Greek alphabet, 815.
Furoic acid, 292.
Fusions: with potassium pyro- .
G1ll¥Detal, 631; analysis of, 633..
,
sulphate, 6.60, 663; with
sodium carbonate, 569-61. Halides, D. in admixture: by ad-
sorption indicators (v), 317;
Gas analysis: by explosion and by indirect method (v). 318~
combustion methods, 767-8; Hardness of water. 466-7; D. of
exercises in, 770 ~ general dis- permane!1t, 466, 471; D. of
cussion of, 732; gr~vimetric temporary, 466, 471; D. of
methods of, 732-; phvsical total. 469; table of, 470.
methods of, 733; titration Heating apparatus, 231.
methods of, 732; volumetric Hehner cylinders, 683.
methods of, 732, 771. Hempel's apparatus, 740; absox:p-
Gas measuring 'f'esseIs. calibration tion pipette for fuming sul-
of, 739. . phuric acid, 743; burette, 745;
Gas volumes. automatic correction double absorption pipette, 742-
to N.T.P., 737; correction for • 3; explosion pipette, 744, 767 ;
temperature and pressure, 736. modified burette, 748; pipette
Index 849
for solid reagents, 742; simple oxidation-reduction reactions,
gas pipette, 741. 112-5; in precipitation reac-
Hempel-Winckler gas burette, 740. tions, 90-95; internal, 365;
Hexamethylenetetramine, 542. ionisation constants of, 59, 61 ;
Hot plates, 231. mixed, 68; neutralisation, 58 ;
Hydrazine, D. of, 447; hydrate, 54~. preparation of solutions of, 67 ;
Hydrochloric acid: composition of screened, 69, 282; theory Clf
constant b.p., 279; Hel con- action of, 58-64; universal or
tent of concentrated acid (v), multiple range, 70.
321; preparation of ~stant International prototype kilogram,
b.p., 279; preparation of O.lN 198.
281; preparation of O.IN Iodates, D. of: as silver iodide (g).
from constant b.p. acid, 280; 582.. . •
standardisation of, by anhy- Iodides, D. of: as palladous iodide
drous sodium carbonate (v), (go), 577; as silver iodide (g),
287, by borax (v), 284, as silver 576 ; by eerie sulphate. (v),
chloride (g), 478. 447; by potassium iodate (v),
Hydrogen ion concentrati.on~ 28; in 441 ; by silver nitrate (v), 316 ;
buffer solutions, 38; in hydro- by Volhard's method (v), 322.
lysed salts, 35-7, 83-4 (see PH). Iodimetry and iodometry, 401 ; de-
Hydrogen ion exponent, 28. tection of end points in, 403 ;
Hydrogen peroxide, D. of: by general discussion of, 401.
manganic sulphate (v), 388; Iodine, puriflcation by sublimation,
by nitrometer or gas volu- . 230.
meter (gv), 774-5; by potas- Iodine solution: oxidising proper-
sium iodate (v), 446; by ties of, 414; preparation of
potassium permanganate (v), O.IN, 415 ; standardisation of,
354; by sodium thiosulphate ' by arsenious oxide, 415, by
(v),425. sodium thiosulphate, 417.
Hydrogen sulphide: D. of (v), 434~. Ion-electron method, 48-51.
precipitations with, 152; re- Ionic product of water, 26-8;
ductions with, 350. strength, 8. .
Hydrolysed salts, titration of, 82. Ionisation constants of acids, 5, 7 ;
lbdrolysis constant, 33; degree of, of bases, 7.
33; of salts, 33. Iron, D. of: as ferric oxide (g), 484 ;
Hydroxides, precipitation and separ- by eupferron (g), 543-4; by
ation of, 155. hexamethylenetetramine (g),
Hydroxylamine, D. of: (v), 458. 542-3; by hydrazine hydrate
p-Hydroxyphenylarsonic acid, 169. (g), 542-4; by potassium
8-Hydroxy-quinoline, 161 (see. cyanate (g), 542-3; by thio-
Oxine). cyanate (c), 701-2; by thio-
Hypoch1orites, D. of: (v), 428. glycollic acid (e), 702-3; in
Hypophosphites, D. of: as magne- brass, 629-30.; in felspar. 663 ;
sium pyrophosphate (g), 587; in German silver, 635; in
as mercurous chloride (g), 587. limestone or dolomite, 659; in
Hyposnlphites, D. of: by thiocyan- portland cement, 668.i total in
ate (v), 322-3. iron ore;by eerie sulphate (v),
383; by potassium dichromate
Ignition of precipitates, 270-4. (v), 369; by potassium per-
Indicators, 43; colour change manganate (v), 352; by titan-
interval of, 58; colour changes ous sulphate (v), 392.
and pH ranges of, 64, 66; Iron and titanium in admixture, D.
external, 116, 365; in neutra- of, 398; ware, 222; wire
lisation reactions, 73, 84-6; in (A.R.), 338.
850 Index
Isopropyl ether, extraction with, aluminium alloys, 653; in
lS3. felspar, 663; in limestone or
LS.W.G., inches and mlllJ., 815. doloI1lite, 658, 661 ; in portland
cement, 668,
Jones' reductor: applications and Main Smith crucible,,272.
limitations of, 347; prepara- Manganese, D. of: as MnNH ,PO"
tion of, 346. H 20 (g), 534; as pyrophos-
Journals of analytical chemistry, phate (g), 533-4; by bismuth-
7S7. ate method (v), 358; by
P~ttinson's method (v), 373; by
Kilogram, 199, S13. )'''' periodate (c), 704; in steel (c),
Kte!dahl's method for nitrOgen, 209. 704; in steel (v), 643.
Manganese dioxide in pyrolusite, D.
Lambert's laws, 671. # of : by nitrometer or gas volu-
Lawrence Smith method, 661. meter (gv), 775'; by potassium
Lead amalgam, preparation .oi, 39S. iodate (v), 445; by potassium
Lead, D. of: as chromate (g), 501- permanganate (v), 355; by
2; as dioxide (e), 501, 610; as sodium thiosulphate· (v), 421.
molybdate (g), 501-2; as Manganic sulphate solution: oxi-
sulphate (g),' 501; by sodium dising properties of, 3S7 ~ pre-
thiosulphate (v), 425; dithi- paration of, 387; standard-
zone method (cl. 725; in isation of, 38,8. >
brass, 627-8; in German silver, Mannitol, 305.
634; sulphide method (cl, 724. Mass action, law of, 4; application
Lead dioxide. analysis of (v), 637. to weak electrolytes, 5.
Liebig's method for cyanide, 327. Mean deviation, 186; relative, 186.
Limestone or dolomite, analysis of, Measuring cylinders, 254; flasks
654. (s~ Volumetrio flasks), 240.
Liquid capacity, measures of: Meniscus correction, 739; double,
apothecaries' (British), S14; for water and mercury, 740.
apothecaries' (U.S.A.), S14; Meniscus, reading position of, 248.
>Imperial or British, S13; U.S., Mercuric oxide, 291.
814. Mercurous nitrate: reagent, 541;
Literature of analytical chemistry, standard solution of, 314.
783-7. . Mercurous perchlorate, standard
Lithium, D. of: as sulphate (g), solution of, 315.
574; in admixture 'with potas- Mercury: cathode, 602; density at
sium.and sodium, 570, 574. various temperatures, 736;
Litmus, 67. poisonous character of, 735;
Litre, 813. purification of, 734.
Logarithms: five figure, 824; four Mercury, D. of: as mercury zinc
figure, 822. thiocyanate (g), 504-5; as
Lovibond all-purpose comparator, sulphide (g), 504; by ethy-
695. lenediamine (g), 504-5; by
Lubricants for glass stop.cocks, 249. potassium iodate (v), 443; in
Lunge's gas volumeter, '737 ; aq:la regia solution, 506.
nitrometer, 771. Meta-cresol purple, 68.
Methyl orange, 67, 450; indigo
Macerated tilte~ paper, 262. carmine, 282.
Magnesia mixture, 517. Methyl red, 67.
Magnesium, D. of: as MgNH ,PO" Methyl yellow, 68.
6HP (g), 494; as pyrophos-' Metric and other units, comparison
phate (g), 494; by oxine (g), of, 813-4.
567; by oxine (v), 454; in Millilitre, 238, 813.
Index 851
Mohr titration: application.of, 311 ; (v). 384; by chloramine-T (v),
theory of, 90. 460 ; by a-naphthyl<tmine-
Molar solutions, 45. sulphanilic acid (c), 700; -by
Molybdate reagent, preparation of, potassium permanganate (v),
465. • 357 u
Molybdenum, D. of: as lead molyb- Nitrogen, D. of Kjeldahl's method,
date (g), 522-3; bya-benzoin- 307.
oxime (g~ 522-4; by oxine (v). Nitron, 165~
523; by oxine (g), 623-4; by a-NitrO-,B-naphthol, 164; prepara-
potassium permanganate (v), tion of. 164.
395; cyc1ohexanol method (el, a-Nitroso-,B-napthol. 163.
711; in steel, 645, 712. • p-Nitrophenol, 68.
Moroney evaporator, 232. Normal solutions, 65.
Mortar, percussion, 234. Note book, laboratory, 473.
Mume furn~ce, 231. N-phenylanthranilic acid, 366.
Munroe cnlbible, 266.
Ohm, 172; law of. Pl.
a-Naphtholphthalein, 67. Organic bases, precipitations with,
Neo-cupferron, 160. 157; precipitants, 158.
Nephelometry, 670, 6~2. Orsat-Lunge apparatus, 751; use of,
Nessler tubes, 678. 752-4.
Nesslerimeters: B.D.H. Lovibond, Oxalates, D. of: as calcium car-
682; Fisher, 681. bonate (g), 587; as calcium
Nessler's reagent, 699. oxide (g). 587; as calcium
Neutral red, 68. oxalate monohydrate (g), 587 ;
Neutralisation curves, 70; of a by eerie sulphate (v): 382; by
polybasic acid with a strong manganic sulphate (v), 388;
base, 80; of strong acids and by potassium permanganate
strong bases, 71; of weak acids (v). 340-1.
and strong bases, 74; of weak Oxidants.: equivalents weights of,
bases and strong acids, 77; of 51-3; table of 52:
:weak bases and weak acids, 78. Oxidation, 48; number method, 51.
Neutralisation reactions, 44. Oxidation potentials, 102; calcula-
Nickel, D. of: as metal (e), 545, 612; tion of. 103; t~ble of standard,
by a-benzildioxime (g), 160; 103.
by dimethylglyoxime (c), 713 ; Oxidation-reduction~ 47; celis, 102 ;
by dimethylglyoxime (g), 497 ; curves, 109; indicators. 112;
by oxine (v), 456; by potas- reactions, 45. 47 ; reactions,
sium cyanide (v), 328; by equilibrium constants of, 104;
pyridine (g), 545-6; by pyri- reactions. theory of, 96.
dine-ammonium thiocyanate Oxidising agents, 48; table of, 54.
(v), 324; by salicylaldoxime Oxine, 161, 557; D. of metals with
(g), 545; in German silver, _(v), 452-8; precipitations witb,
635; in nickel steel (g), 498, 162.
643; in nickel steel (v), 330;
in presence of copper (g), 545 ; Palladium, D. of: as palladous
in 'silver coinage alloy, 639. cyanide (g). 529; by dimethyl-
Nickel silver, analysis of, 634. glyoxime (g), 529-30; by a-
Nickel w~e,.222. nitro-,B-naphthol (g). 529-iJO;
Nitrates, D. of: by gas volumeter by a-nitroso-,B-naphthol (g),
(gv), 773; by nitrometer (gv), ,529-30.
772; by nitron (g). 592; in Parallax, errors due to, 248.
Chile saltpetre (v), 306. Parallel determinations, 191
Nitrites, D. 01: by eerie sulphate Peptisation, 143, 145.
852 Index
Perchlorates. D. of: as potassium 744, 753; slow combustion
~chlorate (g), 581; as silver (Ambler), 769; slow combus-
chloride (g), 581. tion (Dennis), 768; slow com-
Perchloric.acid: explosive dangers bustion (improved), 769 ;
of, 589; use in D. of silicates, sta'ndardised, 246; tolerances
589. of, 246; transfer, 243; use of,
Periodates. D. of: as silver iodide 244.
(g), 582. Platinum apparatus, 222; care and
:Percussion mortar, 234. use of, 274; cleaning and
Peroxides. D. of: (gv), 774; by preservation of, 276.
potassium iodate (v), 445. Poggendorff's compensjlotion method,
Persuiphates, D. of: by potassium , 120. •
permanganate (v), 358. POR.29.
PR, 28; D. of by colorimetric Polarisation e.m.f., 174.
methods, 693-4; D. of by Polarographic methods. 132.
potentiometric methods, 118; Policeman. 223. _.
meters, 131. Porcelain apparatus, 221.
-O-Phenanthroline ferrous ion, 113, Porous porcelain filter crucibles, 267.
366,379. Portland cement, analysis of, 666.
Phenolphthalein, 60, 67. Post precipitation, 148.
.Phenol red, 68. Potassium, D. of: as chloroplatinate
Phenyiarsonie acid, 165. (g), 570-1; as K.Na [Co(NO.l']
Phenylthiobydantoic acid, 164. H.O (g), 571, 573 ; as perchlor-
Phosphates, D. of: as magnesium ate (g), 570, 572; as sulphate
pyrophosphate (g), 584-5; as (g), 571; by ceric sulphate (v),
phosphomolybdic anhydride 472; in admixture with sodium
(g), 584; .(v), 464-5. (g), 570: 573; in felspar, 664.
Phosphites. D. of: as magnesium Potassium bi-iodate, preparation of,
pyrophosphate (g), 586-7; as 293.
mercurous chloride (g), 586. Potassium hi-tartrate, 294.
Phosphor b:conze. 631; analysis of, Potassium bromate solution: oxi-
633. disiug properties of, 460; pre-
Phosphoric acid: action on indi- paration of 0.1N, 4!il.
cators, 82, '115; D. in com- Potassium cyanate. 540.
mercial acid (v), 302; neutral- Potassium cyanide: action on in-
isation of, 8'1. dicators, 83 ; standard solution
Phosphorus, D. of: by potassium of, 329; titration with strong
permanganate (v), 465; by acids, 82.
sodium hydroxide (v), 465; in Potassium dichromate solution:
phosphor bronze, 633 ; in steel, indicators for, 365; oxidising
640. properties of, 364; preparation
Phosphorus, in gas analysis, 764. of O.lN, 366; purification of,
Photo-electric photometer method, 366-7 ; reduction of, 540 ;
677, 689. standardisation of, by iron,
'photometers: gradation, 670; Pul- 367.
frich, 688. Potassium ierricyanide, as external
Phoiometric chemical analysis, 670 .. indicator, 3.65.
Pipettes: absorption (Shepherd), Potassium fiuo-titanate, 706.
754 ; calibration of, 244 ; Potassium hydrogen' phthalate.
+fractional combustion, 753 ; as standard substance, 291:>.
graduated, 243 ; Hempel Potassium iodate solution: for
explosion, 744, 767; in glj,S standardisation of acids, 405;
analysis, 740-4, 754; Lunge- oxidising properties of, 438;
Rey, 263; slow combustion, preparation of 0.025M, 440.
Index 853
Potassium iodide-starch paper, Quartz filter crucibles, 267.
preparation of, 427. Quinaldinic acid: as pH indica.tor,
Potassium llermanganate solution: 68; as precipitant, 166, 510;
application in alkaline solu- recovery of, 167.
tions, 336 ; oxidising pro-
perties of, 334; preparation of Radiator, 232.
O.IN, 338; permanence of, Raman spectra methods. 135.
343; reduction of, 554; Reagents: analytical, 227; quality
standardisation of, by arsen- of, 227; table of analytical,
ious oxide, 339, by ferrous 792-807.
ammonium sulphate, 342, by Recording 01 results, 473.
metallic iron, 341, by oxalic Recrystallisation. purification of
acid or potassium tetroxalate, solids by, 228.
343, by sodium oxalate, 340. Red lead, analysis of, 356.
Potassium tetroxalate, 53, 293. Reducing agents, 48; table of, 54.
Potassium titanium oxalate, 706. Reductants: equivalent weights of,
Potential indicators. 115. 51; table of 52.
Potentials: electrode, 98; stand- Reduction, 48; by amalgamated
ard or normal, 98, 100; stand- zinc, 393; by chromous salts,
ard oxidation, 102, 103. 399; by liquid amalgams, 393 ;
Potentiometer. 121; thermionic by titanous salts, 389; by
valve, 131. vanadous salts, 400; of ferric
Potentiometric titration, 116-7 ; ex- salts, 344; of molybdenum,
perimental details of, 127; 395; of titanium, 394, 396; of
recent developments in, 129; tungsten, 398; of uranium,
theory of, 109. 394; of vanadium, 395.
Precipitate: drying of, 270; igni- Reduction of ferric salts: by hydro-
tion of, 270-4; purity of,147 ; gen sulphide, 350; by Jones'
washing of, 260, 267. reductor, 346; by liquid amal-
Precipitation. 258 ; and super- gams, 393; by stannous chlor-
saturation, 145; completeness ide, 345; by sulphurous acid,
of, 17; conditions of, 149; 349; by zinc and sulphuric
discussion of, 258; fractional, acid, 350.
21, 152; methods, 136; of Relative mean deviation, 186.
hydroxides at controlled pH, Rest point, 201.
155 ; of sulphides, 152 ; Rbodizonic aeid, sodium salt, 332.
practical points about, 258; Rider, 196.
reactions, 45, :no; reactions,
theory of, 86. Salieylaldoxime, 162.
Precision, definition of, 186. Salt effect, 20.
Preventive solution, 345. Salts, 3.
Primary standard substances, 56. Samples: analysed, 792-8; crush-
Probable .error, 188. ing and grinding of, 234;
Probability curve, 188. solution of, 235; weighing oI,
n-Propy!arsonic acid, 649. 205.
,Protective colloid, 144. Sampling of solids, 233; of gases,
Pulfrich photometer. 687; use of, 733-4.
688. Screens. perforated, for crucibles,
Purification 01 substances, 228. 277.
Pyridine. 166. Sectrometer. 131.
Pyrites, D. of sulphur in (g), 482. Selenium, D. of: as element (g),
Pyrogallol, 167; in gas analysis, 525-6; by potassium perman-
764. ga'nate (v), 362; by sodium
Pyrolusite, see Manganese dioxide. thiosulphate (v), 362.
854 Index
Selenium and tellurium in adm.ix- Sodium hydroxide solution: pre--
ture, D. of, 527. paration of carbonate-free, 286-
Separations : by miscellaneous 9: preparation of O.IN, 289;
physical methods, 181; by standardisation of, by acid,
precipitation methods, 152; 289, by potassium hydrogen
electrolytic, 179. . phthalate, 290.
Shepherd apparatus, 754. Sodium thiosulphate solution: pre-
Significant figures, 192. paration of O.lN, 405 ;
Silica apparatus, 222. stability of, 406; standardiza-
Silica, D. of: by molybdate (c), tion of, by copper, 411, by
729 ; in aluminium alloys, potassium broma.te, 408, by
654 ; in felspar. 662; in potassium dichromate, 40!J, by
limestone or dolomite, 655; in potassium iodate, 407, by
insoluble silicates (g), 590 ; potassium permanganate, 409,
in portland cement, 667; in by pure iodine, 413, by stand-
soluble silicates (g), 589; in ard iodine solution, 413.
steel, 642. Solder, analysis of, 636-7.
Silver apparatus, 222. Sofnolite. 593.
Silver, D. of: as chloride (g), 543 ; Sol, 142.
as metal (e), 614; by ammon- Solubility: effect of acids upon, 25 ;
ium thiocyanate (v), 319; by effect of solvents upon, 26;
p-dimethylamino-benzal-rhoda- effect of temperature upon, 26.
nine (c), 728; in alloys (v), Solubility product. 13; of metal
320; in silver coinage alloys, sulphides, 16; principle,
638. limitations of, 19.
Silver· chromate, 90. Spectrophotometry, 670; theory of,
Silver coinage allOY, analysis of, 638. 671.
Silver nitrate solution: preparation Spectroscopic methods, 132.
of O.IN. 31O-11; standardisa- .. Speedyvap " beaker cover, 236.
tion of, by sodium chloride. Standard series method. 675, 678.
311-3. standard solutions, 43; preparation
Soap solution, preparation af of,56.
standard, 468. Standard substances for acidimetry
Soda- IiDle. 593. and alkalimetry: adipic acid.
Sodium anthraquinone-,B-sulphon- 292 ; anhydrous sodium car-
ate, 764; bismuthate, 358; bonate, 281; borax, 284; furoic
chloride, ,310; cobaltinitrite acid, 292 ; mercuric oxide, 291 ;
reagent, 573; diethyldithio- oxalic acid, 293; potassi,um
carbamate, 726-7; hyposul- bi-iodate, 292; potassium bi-
phite, 764; oxalate, 291, 336, tartrate, 294; potassium hy-
MO ; peroxide, analysis of, drogen phthalate, 290; potas-
355. sium tetroxalate, 293; sodium
Sodium carbonate: as standard oxalate, 291; succinic acid,
substance, 281; choice of 292; sulpbamic acid, 294.
indicators for, 84; content of Standard temperature, for volu-
washing soda. 295; prepara- metric glassware, 239.
tion of pure, 281; titration Starch solution: preparation of,
with strong acids, 84. 406-7; stability of, 406-7; use
Sodium, D. of: as magnesium of,496.
uranyl acetate (g), 568-9; as Steam baths, 230; ovens, 230.
zinc uranyl acetate (g), 568; Steel, analysis of, 639.
as sulphate (g), 567-8; by Step photometer method, 676.
potassium permanganate (v), Stirring: in electrolytic analysis,
394; in felspar, 664. 604; rods, 223.
Index 855
Stoichiometric point. 58. Thiocyanate and chloride in a.c1mix-
Storage of solutions in volumetric ture, D. of, 331.
ana1ysis, 254. Thiocyanate. D. of: as barium
Strontium. D. of: as oxide (g), sulphate (g), 578; as cuprous
565; as sulphate (g), 564-5. thiocyanate (g), 577; as silver
Sublimation, purification of solids. thiocyanate (g), 577; by mer-
by, 229. curic nitrate (v), 330; by silver
Succinic acid, 292. nitrate (v), 317.
Sulphamic acid, purification of, 294. Thioglycollic acid, 702.
Sulphate, D. of : as barium sulphate Thiosulphates, D. ot: as barium
(g), 478, 582; with sodium sulphate (g), 584; by iodine
rhodizonate (v), 332. (v), 413 (see Sodium thiosul-
Sulphides, D. of: as barium sulphate).
phate (g), 583; by iodine (v), Thorium, D. of: as oxalate (g), 557-
434. 8; as sebacate (g), 558.
Sulphides, precipitation of, 152. Tin. D. of: as dioxide (g), 519-21,
Sulphites. D. of: as barium suI:: by ammonium iodide procedure,
phate (g); by iodine (V), 433. 520, 632, by thiostannate pro-
Sulphonephthaleins, 68. ' cedure, 626; by chloramine-T
Sulphur, D. of: in iron pyrites (g), (v), 461; by cupferron (g),
482; in portland cement, 668 ; 520-1; by iodine (v), 431-2;
in steel, 640. by phenylarsonie acid (g),
Sulphuric acid, D. of: in concen- 520-1; by potassium iodate
trated acid, 297; in fuming (v), 444; in brass, 626; in
acid,298. bronze, 632; in German silver,
Sulphurous acid and sulphiies: D. 634 ; molybdenum blue method
of (v), 433; reductions with, (c), 723.
349. Titanium, D. of: by ferric salts (v),
Supersaturation, 145. 396 ; by hydrogen peroxide
Snspensoids, 142. (c), 705; by oxine (g), 559;
Syate.matic gravimetric analysis, 500. by oxine (v), 456; by para
hydroxyphenylarsonic acid (g),
559, 561; by potassium per-
Tannin, 165. manganate (v), 395; by
Tartrazine, 313. selenious acid (g), 559-60; by
Tellurium, D. of: as dioxide (g), tannin and antipyrine (g),
526; as element (g), 525-6; by 559-60; in east iron, 707; in
eerie sulphate (v), 385; by felspar, 663, 707; in steel, 647.
potassium dichromate (v), 385; Titanous chloride solution, prepara-
by potassium permanganate tion of, 390; sulphate solution,
(v), 385. preparation of, 390; solutions,
Temperature. standard, 239; cor- standardisation of, 391; salts,
rection for 1 litre glass flasks, reducil}g properties of, 389.
239. Titration, 43; conductometric, 44,
Test papers, 226. 132; curves, 73, 74, 76, 79, 82,
Thallium, D. of: as thallous chro- 85, 88, llI, 112; differential,
mate. (g), 563-4; as thallous 127; error, 43; of borax with
cobaltinitrite (g), 563-4; by a strong acid, 83; of potassium
eerie sulphate (v), 385; by cyanide with a strong acid, 82 ;
potassium iodate (v), 447. of sodium carbonate with a
Thermionic titrimeter without strong acid, 84; oxidation-
batteries, 130. reduction, 110-11; potentio-
Thermionic valve potentiometers, metric, 44, 116; precipitation,
131. 87-8.
856 Index
Thymolphthalein,67. Wash bottle, 219-20.
Tropaeolin O. 68; 00, 68. Washing by decantation, 269; of
Tungsten, D. of: by benzidine precipitates, 150, 267.
(g). 575; by cinchonine (g). Water: ammonia-free, 688; con-
575, 646; by oxine (g), 575-6 ; ductivity, 220 ; distilled, 220;
by potassium permanganate weight to give 1 litre at 20°C,
(v). 398; by tannin-antipyrine 2-l2. •
(g). 375; in steel, 646. Water of hydration, D. of (g), (7-l.
Turbidimetry, 670, 692. Weighing: bottles, 205; errors in,
Turbidity and end point, 95. 208; of chemical samples, 205 ;
Tyndall effect, 141. reduction of, to vacuo, 210-2.
Type metal, D. of antimony and Weighing, methods of: Borda's,
tin in (v). 432. 208 ; direct, 201; Gauss's,
208 ; short swings, 204 ;'
Units of area, 813; of length, SI3; swings, 2W.
of mass. S14; of volume, 238. Weights, 198; apothecaries', 814;
813. . avoirdupois, 814; calibration
Uranium, D. of: as ammonium di- (" of. 212; troy, S14.
uranate (g). 556; by oxine· .Wiping, electrification due to, 209.
(g), 556-7; byoxine (v). 556; by: Witt plate, 265.
potassium permanganate (v).
394. X-ray methods, 135.
Uranyl magnesium acetate, 569; Xylene cyanol FF, 114, 376, 379.
zinc acetate. 569.
Urea., D. ot: (gv). 777. Zero point, 201-2.
Zimmermann-Reinhardt's process.
Vanadium, D. of: as mercurous 335, 352; solution, 335.
vanadat!l (g). 555; as silver Zinc amalgam, preparation of. 397.
vanadate (g), 555-6; by hydro- Zinc, D. of: as metal (e), 551, 614 ;
gen peroxide (cl. 708; by as pyrophosphate (g). 549,
manganic sulphate (v). 388; by 551; as sulphide (g), 550-1;
potassium iodate (v). 448; by as ZnNH.PO. (g), 549, 551;
potassium permanganate (v), as ZnNH.PO •• HtO (g), 531;
395; in steel, 645, 708. by ferrocyanide (v), 463; by
Vanadous salts, reduction with, 400. oxine (g). 457; by oxine (v),
Vapour pressure of water, table of, 457; by pyridine (g), 550,
at various temperatures, 737. 553; by pyridine-ammonium
Vinegar, D. of strength of (v), 297. thiocyanate (v), 324; by
Volatilisation methods, 169. quinaldinic acid (g). 550, 552;
Volhard titration: applications of, in aluminium alloys. 653; in
318-22; theory of. 92. brass, 630-1; in German silver,
Volt, 172. 635; in silver coinage alloys,
Volnmetric analysis: classification 639.
of reactions in, 44; general Zinc dust, evaluation of (gv), 776.
discussion of, 43; storage and Zirconium, D. of: as basic selenite
preservation of solution in, (g). 562-3; as 'pyrophosphate'
254; technique of, 238. (g), 561-3; by oxine (g). 458 ;
Volumetric apparatus, 239. by oxine (v), 457; by para-
Volumetric flasks, 240; calibration hydroxyphenylarsonic acid (g),
of, 241; standardised, 243; 562; by n-propylarsonic acid
tolerances of, 243. (g), 562, 650; in steel, 649.

0034 3059a

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

0034 3059a

Transféré par

Droits d'auteur :

Formats disponibles

A TEXT-BOOK OF

ARTHUR L VOGEL, D.Se. (LOND.); D.Le., F.I.C.

With Plates and Diagrams

LONGM~NS: GREEN AND CO.

LONGMANS, GREEN AND CO.

First Published 1939

CODE NUMBER: 85287.

Printed In Great Britain by

to Messrs. Fisher Scientific Co., of Pittsburgh, Pa., for provid-

1,42. THEORY OF OXIDATION-REDUCTION. REACTIONS 9.6.

THEORY OF GRAVIMETRIC ANALYSIS

GENERAL APPARATUS, REAGENTS AND OPERATIONS

TECHNIQUE OF VOLUMETRIC ANALYSIS

TECHNIQUE OF GRAVIMETRIC ANALYSIS

PRECIPITATION ~ND COMPLEX FORMATION REAC1'IONS

OXIDATION-REDUCTION TITRATIONS. OXIDIMETRY AND

OXIDATIONS WITH POTASSIUM PERl'vIANGANATE

OXIDATIONS WITH CERIC SULPHATE

REDUCTIOXS WITH TITANOUS SALTS

REDUCTIONS WITH AMALGAMATED ZINC AND WITH LIQUID

m, 97. Reductions with chromous salts 399

OXIDATION AND REDUCTION PROCESSES INVOLVING lODINE

IODIMETRY AND IODOMETRY

OXIDATIONS WITH POTASSIUM IODATE

OXIDATIONS WITH POTASSIUM BROMATE

OXIDATIONS WITH CHLORAMINE-T

.. MISCELLANEOUS VOLUMETRIC DETERMINATIONS

SIMPLE qRAVIMETRIC DETERM,INATIONS

SYSTEMATIC GRAVIMETRIC ANALYSIS

GAS-VOLUMETRIC METHODS OF ANALYSIS FOR SOLIDS AND

THE THEORETICAL BASIS OF QUANTITATivE

Phosphoric acid also ionises in stages :

trical forces between the ions. Expressions for the varia!ions

solutions, have also been deduced by the Debye-Hiickel-

Concn x 104 A, a K X 105

The mean classicai or Ostwald dissociation constant of"acetic

It 4. Strengths of acids and bases.-It has already been

Acid. Formula K. pK. = - log K.

This is the rigorously correct expression for the law of mass

duced by Lewis and Randall in 1921, and is defined as equal

The activity coefficient depen~s upon the total ionic strength

I Example 4. What effect has the addition of 0·5 gram molecule

[KCl] I[Cl-] X 10_3 I[Ag+] X 10 8 SAgCl= [~g+].[CqXI010

* Activity = activity coefficient X concentration, i.e., a = fr.. The activity

Substance Solubility pS Substance Solubility pS

A few examples may help the reader to fully understand the

Example 5. The solubility of silver chloride is 0·0015 grams per

Example 6. Calculate the solubility product of silver chromate,

= (2 X 7·5 X 10-5)2 X (7·5 X 10-5)

Example '7. The solubility product of magnesium hydroxide is

I, 9. Completeness of precipitation. Quantitative effects

The solubility of the salt is represented by the [A-] which it

00 0.02 0·04 0·06 0'06 0·10 0·12 0·14 0·16

effect depends upon the valency of the ions of the precipitate

For more concentrated solutions (ft°' s > 0'3) a.n additional

It will be clear from the above short discussion that two

[Ag+J X [NH3]1 [Cd++] x rCN-]·

[ Ag +J X [5 •0 3 --J3 = 1.0 lO-18 [Hg++] X [(,;:'-]4

The application of complex ion formation in chemical separa-

where a"" [xl. andj", refer to the activity, concentration and

[H+] _:_ [O~-]' and [OH-] ='[~:r

Temp. (0C) I K .. X 1014 Temp. (0C)